Research advances in targeting claudin 18.2 in pancreatic ductal adenocarcinoma treatment

Abstract

Claudin 18.2 (CLDN18.2) is a tight junction protein expressed only by gastric mucosal epithelial cells with known functions involving the regulation of cell barriers, polarity, and ion permeability. Research in cancers of the digestive system has shown that CLDN18.2 is often ectopically activated and highly expressed in gastric/gastroesophageal junction adenocarcinoma, and pancreatic cancer, identifying CLDN18.2 as a potential therapeutic target. Immunotherapy strategies targeting CLDN18.2 include monoclonal antibodies, chimeric antigen receptor T cells, diclonal antibodies, and antibody conjugates, and such therapies have shown promising efficacy in gastric cancer. Zolbetuximab, a monoclonal antibody, has also shown promise in the treatment of pancreatic ductal adenocarcinoma (PDAC). This review summarizes and discusses major advances related to the function and targeting of CLDN18.2 in PDAC. By understanding the structure, function, and expression of CLDN18.2 in PDAC as well as the efficacy and mechanisms of treatment methods related to CLDN18.2, we hope to provide insights for the future development of treatment strategies for CLDN18.2-positive PDAC. However, current evidence on the clinical efficacy of targeting CLDN18.2 in PDAC remains preliminary and fragmented, derived mostly from small studies or gastric cancer extrapolation. Much more research is needed before clinical application of CLDN18.2-directed strategies.

Key Words: Claudin 18.2; Pancreatic ductal adenocarcinoma; Targeted therapy; Research progress; Treatment

Core Tip: Claudin 18.2 (CLDN18.2) is expressed in pancreatic ductal adenocarcinoma (PDAC) and represents a promising therapeutic target. This review summarizes current evidence on CLDN18.2-directed therapies, including monoclonal antibodies, antibody drug conjugates, chimeric antigen receptor-T cells, bispecific antibodies, and other emerging strategies. We also outline unique challenges specific to CLDN18.2-positive PDAC, such as an abundance of cancer-associated fibroblasts, a profoundly immunosuppressive microenvironment, and intratumoral antigen heterogeneity. At present, this field remains in its preliminary stage, with most available data derived from small-scale clinical studies or extrapolated from gastric cancer research. Further PDAC-tailored preclinical and clinical investigations are urgently needed.

INTRODUCTION

Pancreatic cancer has been referred to as the “king of cancers”[1] due to its difficult characteristics of insidious onset, strong aggressiveness, poor prognosis, and high mortality rate[2]. Pancreatic ductal adenocarcinoma (PDAC) is the most common type of pancreatic cancer, and with a mortality rate nearly matching its morbidity rate[2]. PDAC specifically has one of the worst prognoses among malignant tumors[3]. Radical surgery is currently the only cure for pancreatic cancer, but more than 80% of patients are unable to receive surgical treatment due to late diagnoses. Systemic therapy is the main treatment method for PDAC discovered at later stages and is mainly based on chemotherapy regimens such as gemcitabine and FOLFIRINOX. Unfortunately, the effectiveness of targeted therapies has been very limited for PDAC. Commonly used drugs such as those targeting the KRAS^G12C and KRAS^G12D genes have shown good efficacy for small percentages of cases with relevant mutations, but they are not effective for most patients with other gene mutations[4-6]. Thus, additional treatment strategies for PDAC are still urgently needed. However, it must be acknowledged that the current evidence base for any novel target in PDAC remains preliminary, and the field is still in an early exploratory stage. Toward this goal, multiple studies in recent years have reported discoveries of important genes and pathways in PDAC, providing new ideas for exploring the pathogenesis of pancreatic cancer and investigating new targets.

The tight junction protein claudin 18.2 (CLDN18.2) has been a focus of research aiming to develop anti-tumor drugs in recent years, as it offers a molecular target with high selectivity and stable overexpression in several malignant tumors, specifically those of the digestive system tumors and their metastases. CLDN18.2 is expressed in up to 70% of PDAC cases but not in normal pancreatic duct and acinar cells, and this high specificity suggests its potential effectiveness as a target in the treatment of CLDN18.2-positive pancreatic cancer. However, several critical caveats should be considered: Reported expression rates vary widely across studies; The 70% figure originates from a study including many precursor lesions [pancreatic intraepithelial neoplasia (PanIN), intraductal papillary mucinous neoplasms, mucinous cystic neoplasms] with near-universal CLDN18 expression; The biological function of CLDN18.2 in PDAC remains unknown; and most evidence comes from preclinical or small-sample studies without PDAC-specific cohorts[7]. The present review summarizes research findings regarding the structure, expression, and mechanisms of action of CLDN18.2 in PDAC and describes recent progress in the development of various therapies targeting CLDN18.2, with the goal of providing a reference for future research related to PDAC treatment. Importantly, this review explicitly recognizes that substantial further basic, translational, and clinical research is still needed before CLDN18.2-directed strategies in PDAC can be considered well established.

STRUCTURE, FUNCTION, AND EXPRESSION OF CLDN18.2

Proteins in the CLDN family are important structural components of tight cell junctions. The 27 isoforms identified to date all have four transmembrane domains, including two extracellular rings, one intracellular ring, and two intracellular short peptides[8]. CLDN proteins connect to the cytoskeleton via zonula occludens-1 and function to maintain tight junction stability and regulate cell polarization and signal transduction[9]. In cancer metastasis, epithelial-mesenchymal transition (EMT) disrupts epithelial junctions, leading to loss of polarity, cytoskeletal reorganization, and cell loss. CLDN proteins provide structural support for actin and signaling proteins in adhesion and migration[10], and thus, alterations in CLDN proteins affect tumor proliferation, invasion, and migration. Located on human chromosome 3q22, the CLDN18 gene is approximately 35 kb in length, consists of 6 exons and 5 introns, and is divided into two isomers, CLDN18.1 and CLDN18.2[11]. Although both isoforms are generated from the same gene via alternative promoters and alternative promoter usage, they exhibit critical differences in the N-terminus and eight amino acids within the first extracellular loop (ECL1)[12]. In terms of gene regulation, CLDN18.2 utilizes exon 1b, which is controlled by the upstream promoter P1 and transcription factor cyclic adenosine monophosphate (cAMP)-responsive element-binding protein (CREB), and is specifically expressed in differentiated gastric mucosal epithelial cells. In contrast, CLDN18.1 employs exon 1a and is regulated by the downstream promoter P2 and transcription factors thyroid-specific enhancer-binding protein/NK2 homeobox 1[13]. CLDN18.1 is selectively expressed in pulmonary alveolar epithelial cells and serves as the predominant isoform in both normal and malignant lung tissues, rendering it an unsuitable drug target. Through gene structure characterization, multi-tissue expression profiling, and cross-species conservation analysis, CLDN18.2 has been confirmed to exhibit stringent lineage- and tissue-specific expression in differentiated gastric mucosal epithelial cells, positioning it as a highly promising target for tumor-targeted therapy[11]. CLDN18.2 is mainly involved in regulating the cell permeability to cations such as hydrogen ion and sodion, and deletion of CLDN18.2 can trigger inflammatory responses and neutrophil aggregation[14]. During tumorigenesis, loss of cell polarity results in exposure of epitopes of this protein that are normally masked by tight junctions, resulting in its ectopic expression. This is commonly observed in PDAC, gastric/gastroesophageal junction (G/GEJ) adenocarcinoma, ovarian cancer, and others[12]. CLDN18.2 expression exhibits marked heterogeneity, with significant differences in expression levels observed even among different regions within the same tumor[15]. Moreover, its expression level often declines with tumor dedifferentiation, with the greatest reductions often observed in poorly differentiated and undifferentiated carcinomas[16]. Paired analyses have shown that CLDN18.2 expression in metastatic lesions can be lost, gained, or remain highly concordant compared with the primary tumor[15,17], suggesting that biopsy evaluation of metastatic lesions is warranted when feasible. Furthermore, recent studies indicate that CLDN18.2 expression can undergo dynamic changes during disease progression, especially downregulation under the selective pressure of chemotherapy[18-20], highlighting the importance of biopsy timing for companion diagnostics. In summary, research to date indicates the pathological potential of CLDN18.2 as a therapeutic target in multiple solid tumors, including pancreatic cancer. Its expression represents a druggable target, and clinical accessibility can be achieved through rational patient selection strategies.

SIGNALING PATHWAYS OF CLDN18.2

The activation of CLDN18.2 is known to be regulated by the transcription factor CREB[11,21]. An investigation of the molecular structure and tissue distribution of CLDN18.2 in various mammals demonstrated that the protein sequence of CLDN18.2 and the promoter regions of its orthologs, including the binding sites for CREB, exhibit high conservation and homology, particularly in the relevant domains of the CLDN18.2 proteins of mice, rats, rabbits, dogs, monkeys, humans, and lizards. These findings reveal the stability of CLDN18.2 in evolution and provide an important theoretical basis for the development of broad-spectrum anti-tumor drugs[11]. However, CREB phosphorylation cannot be employed as the only mechanism for activation of CLDN18.2. A genome-wide analysis by Zhang et al[21] demonstrated that activation of CLDN18.2 depends on the binding of the transcription factor CREB to hypomethylated CpG islands. Specifically, when cAMP in tumor cells activates protein kinase A (PKA), the activated PKA enters the nucleus and phosphorylates the Ser133 site of CREB (a key step in CREB activation). However, whether CREB can bind to the promoter of CLDN18.2 depends on the presence of a cAMP-responsive element (CRE) near the promoter as well as the methylation status of the CRE, with methylation of the CRE in the CpG island of the promoter region significantly inhibiting CREB binding. Therefore, the promoter region of CLDN18.2 must contain an unmethylated functional CRE to serve as a binding target for CREB (Figure 1). In addition, expression of only a few target genes will be induced after extensive phosphorylation, primarily due to the selective recruitment of co-activators. Phosphorylated CREB must function in concert with the co-activator CREB binding protein to recruit transcription factors to its promoter region to drive abnormal activation and overexpression of CLDN18.2[21].

Figure 1 Signaling pathway of claudin 18.2. When the generated second messenger cyclic adenosine monophosphate activates protein kinase A (PKA) in tumor cells, the activated PKA enters the nucleus and phosphorylates cyclic adenosine monophosphate-responsive element-binding protein (CREB). When the claudin 18.2 promoter region has an unmethylated functional CRE, binding to CREB can occur to drive claudin 18.2 overexpression. CLDN18.2: Claudin 18.2; cAMP: Cyclic adenosine monophosphate; ATP: Adenosine triphosphate; CRE: Cyclic adenosine monophosphate-responsive element; CREB: Cyclic adenosine monophosphate-responsive element-binding protein.

Ito et al[22] studied the effect of protein kinase C (PKC) activator (TPA) on CLDN18 expression in pancreatic cancer using four pancreatic cancer cell lines (HPAF-II, HPAC, PANC-1, and BXPC3) as well as human telomerase reverse transcriptase (hTERT)-expressing human pancreatic duct epithelial (HPDE) cells (hTERT-HPDE cells) obtained by introducing the hTERT gene into primary cultured HPDE cells. They found that CLDN18 messenger RNA expression was significantly induced by TPA in all four human pancreatic cancer cell lines and in hTERT-HPDE cells, and that CLDN18 protein expression also was significantly increased in the highly differentiated or medium-differentiated human pancreatic cancer cells of the HPAF-II and HPAC lines and in hTERT-HPDE cells. Moreover, in the human pancreatic cancer cell lines, PKCδ, PKCε, and PKCα inhibitors blocked the upregulation of Cldn18 by TPA, whereas in hTERT-HPDE cells, the TPA-mediated upregulation of Cldn18 was blocked by PKCδ, PKCθ, and PKCα inhibitors. In addition, they identified a CpG island in the coding sequence of Cldn18, and treatment with demethylating agents enhanced the upregulating effect of TPA on Cldn18 in HPAF-II and HPAC cells, but not in hTERT-HPDE cells. The findings of their research suggest that CLDN18 expression is mainly regulated at the transcriptional level through specific PKC signaling pathways in human pancreatic cancer cells and is affected by changes in DNA methylation.

EXPRESSION OF CLDN18.2 IN PDAC

Upregulated expression of CLDN18.2 has been demonstrated in digestive malignancies, including esophageal cancer and pancreatic cancer[23]. Because CLDN18.2 is not expressed in normal pancreatic ductal and acinar cells, this expression difference suggests that it may be a useful pathological diagnostic marker for PDAC. Wöll et al[24] found that among 202 cases of primary PDAC, 59.2% were positive for CLDN18.2, and 54.6% of cases showed a CLDN18.2 staining intensity > 2+. Among patients with neuroendocrine neoplasia in their study, 20% of cases were positive for CLDN18.2 expression, and the staining intensity for CLDN18.2 was ≥ 2+ in all of these cases. Furthermore, the positivity rates for CLDN18.2 expression in pancreatic cancer cases with lymph node metastasis, liver metastasis, and no metastasis group were 70%, 66%, and 59%, respectively. These results indicate that CLDN18.2 expression is not only highly prevalent in primary pancreatic cancer but also in metastatic lesions, and thus, a considerable number of pancreatic cancer patients might benefit from CLDN18.2-based targeted treatments. Lyu et al[25] investigated CLDN18.2 expression in 309 patients with PDAC who underwent curative pancreatectomy, and through antibody staining of surgical specimens observed positive staining for CLDN18.2 in 94 (30.4%) patients (staining intensity ≥ 2+ in 75% of tumor cells). They further reported that CLDN18.2-positive cases showed higher PDAC differentiation and a median overall survival (OS) of 30 months, which was significantly longer than that of CLDN18.2-negative cases (18 months). Additionally, the OS rate of patients with CLDN18.2-positive PDAC was higher. Correlation analysis showed that in PDAC patients with lymph node metastasis and nerve invasion, CLDN18.2 expression was significantly increased, and patients with CLDN18.2-positive PDAC had a better prognosis than those with CLDN18.2-negative PDAC. Their analysis also identified lymph node metastasis as an independent factor negatively correlated with the prognosis of PDAC patients, suggesting that CLDN18.2 could be used as a predictor of prognosis for PDAC patients[25].

Zhang et al[26] examined the expression pattern of CLDN18.2 in PDAC and its relationship with EMT phenotype using a microarray cohort of 302 PDAC specimens and a whole-slide cohort of 84 PDAC specimens. They found that CLDN18.2 was more commonly expressed in highly differentiated PDAC (82%) and intermediate-differentiated PDAC (63%) compared with low differentiated PDAC (10%), and statistical analysis confirmed that CLDN18.2 positivity was significantly correlated with the degree of pathological differentiation of PDAC (P < 0.001). CLDN18.2 positivity was also more common in women and non-smokers, and CLDN18.2-positive patients were more likely to have abdominal pain and jaundice but less likely to have common bile duct invasion and distant metastasis. In their study, CLDN18.2 expression in PDAC was associated with improved OS but not with progression-free survival (PFS) or patient prognosis[26]. They also examined the expression of the EMT marker vimentin in the PDAC cases. Of six cases that were positive for vimentin, none expressed CLDN18.2, suggesting that the intercellular junction was destroyed during EMT, and the loss of CLDN18.2 expression was related to tumor invasion and metastasis. To analyze the expression of CLDN18.2 in precancerous lesions, they also examined 10 cases of PDAC with (PanIN). All 10 PanIN cases were CLDN18.2 positive, while adjacent normal ductal epithelium was negative, suggesting that CLDN18.2 may have been abnormally expressed in the early stage of cancer development and may be involved in the transition process from normal epithelium to precancerous lesions, making it a potential marker for early diagnosis of PDAC or precancerous disease monitoring[26]. Another study investigated CLDN18.2 expression by immunohistochemistry (IHC) in 68 PDAC cases who underwent surgery, and 54.4% (n = 37) of the patient samples were positive for CLDN18.2 expression, with staining intensity scores of + 1 in 30.8% (n = 21), + 2 in 16.2% (n = 11), and + 3 in 7.4% (n = 5). The OS of cases with high CLDN18.2 expression was longer than that of patients with lower CLDN18.2 expression, which also suggests that CLDN18.2 may be a useful prognostic marker for PDAC[27]. The key findings of the above studies on CLDN18.2 expression in PDAC are summarized in Table 1[24-27].

Table 1 Summary of pancreatic cancer research involving analysis of claudin 18.2 expression.

Experimental goal	Number/type of cases	Core results	Conclusions	Ref.
Study the expression of CLDN18.2 in normal pancreatic tissue, primary lesions, metastases, and intra-individual matched samples in patients with PDAC, NEN, and acinar cell carcinoma	24 cases of normal pancreatic tissue, 202 samples of primary lesions, 84 cases with metastasis, and 48 intra-individual matched samples	Positive rate: 59.2% in primary PDAC. 20% in NEN. 69.4% in lymph node metastasis. 65.7% in liver metastasis	In principle, many pancreatic cancer patients can benefit from CLDN18.2-based targeted therapy	Wöll et al[24]
Study on the expression of CLDN18.2 in PDAC paraffin specimens and prognosis in patients undergoing curative pancreatectomy	309 cases of PDAC treated with curative pancreatectomy	30.4% of patients had ≥ 75% of tumor cells with a staining intensity of ≥ 2+; median OS of patients with CLDN18.2 positivity (30 months) was significantly longer than that of patients with CLDN18.2 negativity (18 months)	In PDAC patients with lymph node metastasis and perineural invasion, CLDN18.2 expression can be used as a prognostic indicator	Lyu et al[25]
Study on the relationship between CLDN18.2 expression and EMT using PDAC tissue microarrays and whole sections	TMA cohort: 302 cases. Whole slide cohort: 84 cases. Precancerous lesion analysis: 10 cases with PanIN	The CLDN18.2 positivity rate correlated significantly with the degree of pathological differentiation of PDAC. None of the 6 vimentin-positive cases expressed CLDN18.2. All 10 cases of PanIN were CLDN18.2 positive, while the adjacent normal ductal epithelium was negative	Loss of CLDN18.2 expression is related to tumor invasion and metastasis. CLDN18.2 may be involved in early carcinogenic transformation and can be used as a marker for early diagnosis of PDAC or monitoring of precancerous lesions	Zhang et al[26]
Study on the correlation between CLDN18.2 expression and survival in patients with resectable PDAC	68 PDAC patients who underwent surgery	54.4% of cases were CLDN18.2 positive (30.8% with IHC staining intensity of 1+, 16.2% with 2+, and 7.4% with 3+)	CLDN18.2 can serve as a prognostic marker for PDAC	Kayikcioglu and Yüceer[27]

CLDN18.2: Claudin 18.2; PDAC: Pancreatic ductal adenocarcinoma; NEN: Neuroendocrine neoplasia; EMT: Epithelial-mesenchymal transition; TMA: Tissue microarray; PanIN: Pancreatic intraepithelial neoplasia; IHC: Immunohistochemistry; OS: Overall survival.

DIAGNOSIS AND TREATMENT OF CLDN18.2-POSITIVE PANCREATIC CANCER

Diagnosis of CLDN18.2-positive tumors

IHC is the primary method for determining CLDN18.2 expression in tissues, either qualitatively and quantitatively, and is suitable for medical research and clinical diagnostic use[28]. In October 2024, the VENTANA CLDN18 (43-14A) RxDx test became the first IHC companion diagnostic test to receive the European Conformity mark approval as well as approval by the United States Food and Drug Administration (FDA)[29]. Another valuable method for detecting CLDN18.2 expression is molecular targeted probe technology, which uses “molecular probes” to specifically identify target molecules, which can be proteins, nucleic acids, small molecule metabolites, etc., and achieve localization, quantification, or dynamic tracking of target molecules through signal markers carried by the probes. Previous studies have developed molecular imaging strategies based on a specific antibody, 5C9, for CLDN18.2. The synthetic ¹²⁴I-5C9 positron emission tomography (PET) probe and the cyanine 5.5-5C9 fluorescent probe were experimentally verified to be specific for CLDN18.2 and to effectively detect CLDN18.2-positive tumors in xenograft models. FD1080-5C9[30] is a near-infrared fluorescence zone II probe that shown to be useful in surgical resection. In an orthotopic tumor model, complete tumor resection could be achieved with the use of the combination of ¹²⁴I-5C9 and FD1080-5C9 indicating these probes may serve as valuable tools for clinical diagnosis and surgical guidance, providing strategies for accurate detection and surgical guidance for CLDN18.2-positive cancers[30]. In addition, a novel ⁶⁸gallium (⁶⁸Ga)-labeled nanobody probe [(⁶⁸Ga) Ga-PMD22] has been developed for PET/computed tomography imaging of CLDN18.2. Owing to its ultra-small size, high penetrability, rapid clearance, and high specificity, it offers improved sensitivity and accuracy for tumor diagnosis, with preliminary clinical data supporting its translational potential[31]. In addition to tissue-based IHC and molecular probe technology, liquid biopsy represents an emerging non-invasive approach for detecting CLDN18.2 expression in PDAC. Although direct evidence in PDAC is still limited, studies in gastric cancer have demonstrated that CLDN18.2-positive exosomes or circulating tumor cells can be isolated from peripheral blood and quantified using flow cytometry or digital polymerase chain reaction[32,33]. Such approaches could potentially overcome the challenges of intratumoral heterogeneity and the difficulty of obtaining serial tissue biopsies in PDAC patients. However, the sensitivity and specificity of CLDN18.2 detection in liquid biopsies remain to be validated, and no standardized protocol has been established for PDAC. Future research should explore whether CLDN18.2 in circulating components correlates with treatment response and prognosis.

Given that CLDN18.2 may be associated with tumor invasion and metastasis, detecting its expression remains relevant, but PDAC-specific challenges (heterogeneity, desmoplastic barrier, biopsy difficulty) preclude direct extrapolation from gastric cancer. Future work requires PDAC-dedicated validation cohorts.

Treatment of CLDN18.2-positive tumors

At present, targeted treatments for CLDN18.2-positive tumors mainly include monoclonal antibodies (mAbs), antibody drug conjugates (ADCs), chimeric antigen receptor T cells (CAR-T cells), and bispecific antibodies (BsAbs) (Figure 2). More than 20 drugs have entered phase III trials, and some have been qualified by the United States FDA. In addition, new technologies are being explored, such as nano-antibodies, genetically engineered membrane-based nano-connectors, and trispecific antibodies, which will likely further expand the boundaries of CLDN18.2-based targeted therapy through multi-dimensional innovation. However, due to the expression of CLDN18.2 in normal gastric mucosal epithelial cells, there is a potential risk of off-target tumor toxicity in the gastric mucosa in targeted therapy, which has been a practical problem observed in clinical practice and is the reason for current the additional use of antiemetic agents in treatment[34]. Furthermore, all therapeutic approaches for CLDN18.2-positive PDAC, including chemotherapy, targeted therapy, immune checkpoint inhibitors, and cellular therapy, are constrained by three common challenges. Firstly, tumor cells can form subclones by downregulating CLDN18.2 or upregulating its variants, such as CLDN18.1. This impairs the ability of CAR-T cells to recognize and eliminate tumor cells, ultimately leading to antigen loss[35,36]. Secondly, in the immunosuppressive tumor microenvironment (TME), various cell types [such as cancer-associated fibroblasts (CAFs), regulatory T cells (Tregs)] and stromal components hinder the function of immune cells through physical barriers and immunosuppressive effects[37-39]. Thirdly, the continuous activation of the transforming growth factor beta (TGF-β) signaling pathway and the upregulation of co-inhibitory signals such as programmed death ligand 1 (PD-L1)/programmed death 1 (PD-1) jointly mediate the decrease in antitumor activity and functional exhaustion of effector T cells, resulting in unsatisfactory durability of various immunotherapies[40-42]. While the three common challenges outlined above are well recognized, their relative contributions to treatment failure in PDAC remain poorly quantified. Moreover, most of the preclinical models used to study these challenges do not fully recapitulate the unique desmoplastic and immunosuppressive features of human PDAC. Strategies to overcome antigen loss, stromal barriers, and T-cell exhaustion have largely been developed in gastric cancer contexts and lack PDAC-specific validation. Therefore, translating these insights into effective CLDN18.2-targeted therapies for PDAC requires dedicated models and biomarker-driven trials.

Figure 2 Targeted treatments for claudin 18.2-positive tumors. The main forms employ monoclonal antibodies, antibody drug conjugates, chimeric antigen receptor T cells, and bispecific antibodies. CLDN18.2: Claudin 18.2; mAb: Monoclonal antibodies; ADC: Antibody drug conjugates; BsAb: Bispecific antibodies; CAR-T: Chimeric antigen receptor T cell; CD3: Cluster of differentiation 3.

mAbs

The probability of CLDN18.2 binding to antibodies in normal tissues is extremely low, but the cell polarity disorder related to oncogenic effects causes the binding site of CLDN18.2 on the cell surface to be exposed, allowing it to be targeted[43]. Several mAbs with antibody-dependent cytotoxicity (ADCC) activity, including AB011[44], ASKB589[45,46], FG-M108[47], MIL93[48], osemitamab[49], FL-301, ZL-1211[50], and zolbetuximab[51], have entered clinical trials for CLDN18.2-based targeted therapy. Of these, phase III clinical trials have been completed in G/GEJ cancer for ASKB589, osemitamab, FG-M108, and zolbetuximab[52]. Zolbetuximab was the first specific mAb approved by the FDA for clinical trial testing, and the first specific mAb designed for the treatment of CLDN18.2-positive advanced G/GEJ cancer, mainly through ADCC and complement-dependent cytotoxicity (CDC) to induce tumor cell lysis[53]. Early research on zolbetuximab confirmed that this mAb can specifically bind to CLDN18.2-positive pancreatic cancer cells, and using 26 human pancreatic cancer cell lines for subcutaneous and intravenous xenotransplantation in nude mice, found that higher cell surface expression of CLDN18.2 corresponded to stronger mAb binding[54]. The ADCC cleavage rate was reported to be > 78% in transduced cells and 48%-68% in endogenous cells, but only transduced cells with high CLDN18.2 expression were susceptible to killing via CDC. In addition, the combination of zolbetuximab and gemcitabine was shown to enhance the ADCC effect, delay tumor growth, prolong survival, and reduce metastasis in a mouse model of pancreatic cancer[54]. For the treatment of PDAC specifically, zolbetuximab has been evaluated in the phase II GLEAM trial; however, the trial did not meet its primary OS and PFS endpoints, and further enrollment for the PDAC arm was discontinued in February 2026[55]; no phase III randomized controlled trial (RCT) for PDAC has been initiated to date. Real-world evidence indicates that approximately 30.4% of PDAC patients are CLDN18.2-positive and suitable for treatment with zolbetuximab, and patients who received zolbetuximab combined with chemotherapy maintained a longer health-related quality of life than those not treated with zolbetuximab[25]. Consistently, in another study examining 120 surgically resected PDAC samples, 32.5% of the samples stained positively for CLDN18.2 using 43-14A antibodies and the criteria of ≥ 75% of tumor cells with ≥ 2+ membrane staining. In that study, CLDN18.2 positivity was significantly associated with a lower tumor grade (P < 0.05) and a significantly higher survival rate at the median OS (23 months). These results further suggest that approximately one-third of PDAC patients are likely to benefit from CLDN18.2-based targeted therapies such as zolbetuximab[56]. The results of another phase I/IIa study of the effectiveness of a combination of FG-M108 and chemotherapy for pancreatic cancer treatment were presented at the 2025 American Society of Clinical Oncology (ASCO) meeting. In that study, patients with CLDN18.2-positive (IHC-defined as ≥ 10% tumor cells with 1+/2+/3+ membranous staining) locally advanced or metastatic pancreatic cancer received FG-M108 plus gemcitabine and nab-paclitaxel. The overall objective response rate (ORR) after treatment with FG-M108 plus gemcitabine and nab-paclitaxel was 32.5%, and the disease control rate (DCR) was 100%, with a median PFS (mPFS) of 6.8 months[57], showing promising antitumor activity in the first-line PDAC setting[58]. Notably, a case of a pretreated patient whose target lesions shrank by 86% at week 6 and achieved a durable complete response (CR) at day 270 was reported at American Association for Cancer Research 2026 (Abstract 2645), representing an encouraging early efficacy signal for this agent in the PDAC setting[59]. To overcome the physical barrier of PDAC stroma, preclinical studies have explored the combination of mAbs with stromal-modulating agents, pegvorhyaluronidase alfa (PEGPH20) enhances T-cell infiltration and drug delivery via hyaluronan degradation. Preclinical studies revealed that PEGPH20 combined with focal adhesion kinase (FAK) inhibitors and anti-PD-1 antibodies increases T-cell infiltration, promotes effector memory T-cell differentiation, and reduces granulocytic myeloid-derived suppressor cells (MDSCs), thereby sensitizing PDAC to immunotherapy[60]. Consistently, the phase II trial PCRT16-001 showed that PEGPH20 plus pembrolizumab achieved a median OS of 7.2 months and mPFS of 1.5 months in pretreated hyaluronan-high metastatic PDAC[61]. Moreover, FAK and proline-rich tyrosine kinase 2 are activated in PDAC tissues, stroma, and cell lines, The FAK inhibitor PF-562271 dose dependently inhibits FAK Y397 phosphorylation, suppresses PDAC growth and metastasis, and remodels the TME[62], though its has not yet been tested specifically with CLDN18.2-targeting antibodies. Despite promising preclinical and early clinical results, limitations exist for zolbetuximab in PDAC. Current evidence is mostly from phase II trials or subgroup analyses, with no dedicated phase III RCT in PDAC having been reported. Variable CLDN18.2 positivity due to different clones and cutoffs limits cross-study comparisons. Combinations of CLDN18.2 antibodies with stromal modulators such as PEGPH20 and FAK inhibitors remain untested clinically.

PDAC-specific milestone summary

Zolbetuximab: The phase II GLEAM trial failed OS/PFS endpoints; no phase III trial has been initiated in PDAC. This first RCT, albeit negative, provides profound lessons for the field. FG-M108: Has received breakthrough therapy designation from China’s National Medical Products Administration for PDAC with phase III clinical trials now initiated. Osemitamab: Currently in phase I/IIa trials and has been granted orphan drug designation. Other mAbs remain in phase I/II trials. Overall: FG-M108 has become the first to enter phase III registrational trials specifically for this indication, marking a critical step toward clinical translation.

ADCs

ADCs are targeted therapy drugs in which mAbs are conjugated to small molecule drugs with strong cytotoxicity through chemical linkers. Accordingly, ADCs can specifically recognize and bind highly expressed antigens on the tumor cell surface, enter the tumor cell to release toxins, and thereby specifically kill tumor cells. Multiple ADCs targeting CLDN18.2 are being tested in clinical trials in China and elsewhere, including ATG-022[63], BL-M05D1[64], AZD0901 (CMG901)[65], EO-3021 (CPO102 or SYSA1801)[66], IBI343[67], JS107, RC118, SHR-A1904[68], SKB315, LM-302[69], SOT102[70], TQB2103, and XNW27011[71]. Among these, BL-M05D1, IBI343, and SOT102 are loaded with the DNA topoisomerase inhibitors alkaloid Ed-04, exatecan, and the anthracycline derivative PNU-159682, respectively. Accordingly, these ADCs can inhibit the proliferation of tumor cells via the destruction of DNA structure, topoisomerase activity, or interference with DNA replication and transcription through the formation of a DNA topoisomerase inhibitor ternary complex[64,67,70]. The ADCs ATG-022, AZD0901, EO-3021, RC118, and LM-302 are prepared by conjugating a tubulin inhibitor [monomethyl auristatin E (MMAE)] with humanized CLDN18.2 mAbs. As such, these ADCs can bind CLDN18.2-positive cells with high affinity, release MMAE through ADC internalization, directly lead to tumor cell cycle arrest and induce apoptosis, activate the immune defense, and kill tumor cells through ADCC and CDC effects[58,65-67]. This ADC type was shown to inhibit the growth of CLDN18.2-positive gastric and pancreatic cancer cells in cell line-derived xenograft (CDX) models[72]. After a CLDN18.2-ADC binds to the extracellular domain of its target, it is internalized to the lysosomal compartment, after which it was found to induce persistent and complete tumor regression in CLDN18.2-positive CDX models and patient-derived xenograft (PDX) models[72]. Currently, the CLDN18.2-targeting ADCs AZD0901, LM-302, and IBI343 that been approved for testing in phase III clinical trials for CLDN18.2-positive G/GEJ cancer[71]. Among these, AZD0901 is the first CLDN18.2 ADC to be tested in clinical trials and also the first and only investigational drug targeting CLDN18.2 to obtain the “fast track” designation from the FDA[73]. A clinical trial targeting the CLDN18.2-specific ADC CMG901 for advanced G/GEJ cancer included 89 evaluable CLDN18.2-positive patients, defined by ≥ 20% of tumor cells with 2+ membrane staining intensity, and reported an ORR of 32.6%, a mPFS of 4.76 months (95% confidence interval: 3.35-6.14 months), and a 9-month OS rate of 56.4%, with a median OS not reached[64]. Meanwhile, a phase II study evaluating AZD0901 in advanced solid tumors, including pancreatic cancer, is ongoing[65]. A clinical study of the effectiveness of LM-302 for the treatment of refractory advanced G/GEJ cancer enrolled patients with ≥ 50% of tumor cells staining positive for CLDN18.2 with an IHC expression level of 2+/3+. In these patients, LM-302 treatment led to an ORR of 30.6%, DCR of 75.0%, and a mPFS of 7.2 months[52]. The IBI343 ADC has been tested for the treatment of refractory advanced PDAC or biliary tract cancer. For patients with a CLDN18.2 positivity rate of ≥ 40% at an IHC expression level of 1+/2+/3+, IBI343 has shown certain therapeutic effects, with an ORR of 28.0% and a DCR of 80.0%; however, PFS data were not yet available[58]. At the 2025 ASCO Annual Meeting, updated results from the phase I dose-expansion cohort of IBI343 were presented. In the 6 mg/kg dose group, 44 patients with CLDN18.2-positive PDAC achieved a confirmed ORR of 22.7%, a DCR of 81.8%, a mPFS of 5.4 months, and a median OS of 8.5 months[74]. Based on these encouraging efficacy signals, IBI343 has now advanced into a phase III registrational trial specifically in PDAC. Overall, ADCs potentially offer new treatment options for patients with CLDN18.2-positive refractory gastrointestinal tumors. In one study, dual-targeted, potential of hydrogen-responsive nanoparticles that co-deliver AV3 and KH3 targeted stromal barriers and metabolism, inhibiting tumor growth and downregulating alpha-smooth muscle actin (α-SMA), collagen 1A1, and PD-L1 in PANC-1 xenografts[75]. Nanoparticle-based delivery systems or conjugation with stroma-penetrating peptides may enhance ADC penetration through the desmoplastic stroma, an area requiring further investigation in PDAC models. While several CLDN18.2-targeting ADCs have reached phase III investigation in gastric cancer, their development in PDAC is still preliminary. Their efficacy relies on CLDN18.2 expression and internalization, but high heterogeneity of CLDN18.2 expression in PDAC may cause primary resistance in many patients. In addition, off-target toxicities such as MMAE-related peripheral neuropathy remain poorly defined in PDAC, and nanoparticle delivery or stroma-penetrating strategies are still preclinical.

PDAC-specific milestone summary

IBI343 has advanced to a PDAC-dedicated phase III registrational trial, supported by a completed phase I expansion cohort in PDAC. BA1301 has received United States FDA orphan drug designation; other ADC candidates are still in early exploration; and SOT102 has been discontinued. Overall, IBI343 represents the most advanced CLDN18.2-targeted ADC among PDACs, while other drugs remain in preclinical or early-stage clinical stages.

CAR-T cell immunotherapy

CAR-T cell immunotherapy has demonstrated favorable therapeutic effects in hematological malignancies; however, progress regarding its use to treat solid tumors has been slow. Currently, CAR-T cell therapies targeting the CLDN18.2 antigen show great potential in various solid tumors, including G/GEJ adenocarcinoma, and pancreatic cancer, but the evidence is mostly based on preclinical studies. CT041 CAR-T is the first-in-class CAR-T therapy for solid tumors to be filed for marketing approval globally[76]. A study of 37 patients (including 5 patients with CLDN18.2-positive PDAC) reported that treatment with CT041 CAR-T cells resulted in an ORR of 48.6%. These results indicate that CT041 CAR-T cells exhibit favorable therapeutic efficacy in CLDN18.2-positive PDAC, supporting its further development as a potential treatment option for this refractory malignancy[77]. Research has demonstrated that the TME of PDAC is rich in CAFs, which restrict the infiltration of CAR-T cells. Accordingly, researchers developed CLDN18.2-targeted CAR-T cells co-expressing CXCR4 (referred to as CXCR4 CAR-T cells), the receptor for CAF-secreted chemokine stromal cell-derived factor 1α (SDF-1α), and tested them in a PDAC mouse model with normal immune function[78]. Mechanistically, CXCR4 inhibits signal transducer and activator of transcription 3 (STAT3)/nuclear factor kappa-B (NF-κB) signaling in CAR-T cells via the STAT3/NF-κB/SDF-1α axis, thereby reducing the secretion of SDF-1α by CAFs, suppressing the recruitment of MDSCs to tumor sites, alleviating immunosuppression, and enhancing anti-tumor efficacy[78]. In another study, CLDN18.2-targeting CAR-T cells expressing truncated epidermal growth factor receptor were used to treat three patients with advanced pancreatic cancer. Among them, one patient achieved a CR and two achieved a partial response (PR), demonstrating strong anti-tumor activity of the tested CAR T-cells against pancreatic cancer[79]. Additionally, in patients with metastatic pancreatic cancer, CLDN18.2-targeting CT041 CAR-T cell therapy enabled patients to achieve a PR or even a CR[80]. Notably, CAR-T cell therapy requires pre-treatment to achieve immune cell depletion, which may lead to complications such as increased infection risk, myelosuppression, and immune deficiency in patients. Research has indicated that the co-expression of interleukin (IL)-7 and chemokine ligand 21 in CAR-T cells (referred to as 721 CAR-T cells) can enhance their anti-tumor activity[81]. These cells have exhibited therapeutic effects in multiple solid tumor models without the need for pre-treatment, inhibiting the growth of tumors with heterogeneous antigen expression and even inducing tumor remission. Moreover, 721 CAR-T cells can increase the infiltration of T cells and dendritic cells and inhibit tumor angiogenesis. Based on these findings, 721 CAR-T cells represent a promising therapeutic strategy for solid tumors[81].

To overcome the common barriers that limit PDAC treatment across different modalities, multiple CAR-T cell engineering strategies have been developed, and preclinical studies have explored the construction of dual-target CAR-T cells against fibroblast activation protein alpha and CLDN18.2[82]. These cells improve the immunosuppressive TME by reducing the recruitment of MDSCs, while decreasing T-cell exhaustion in a TGF-β dependent manner and enhancing the survival and function of cluster of differentiation (CD) 8+ T cells. Additionally, researchers have investigated the ability of ectopic expression of CXCR2 to enhance the tumor-targeted infiltration and retention capabilities of CAR-T cells, reduce MDSC and CXCR2+ macrophage infiltration, lower IL-8 levels, and partially reverse the immunosuppressive microenvironment[83]. Furthermore, these dual-target CAR-T cells were shown to effectively eliminate CAFs, reduce fibrotic barriers, and promote T-cell infiltration. Another study[84] used genetic engineering to enable CAR-T cells to express dominant-negative dominant negative TGF-β receptor type II, which blocks the phosphorylation of SMAD2/SMAD3 and counteracts the effects of TGF-β. In a PDX model of pancreatic cancer, this mechanism significantly enhanced the anti-tumor efficacy of CAR-T cells with an efficacy approximately double that of unmodified CAR-T cells, demonstrating the effectiveness of optimizing CAR-T therapy for solid tumors by counteracting inhibitory signals in the TME[84]. In another approach, adoption of the short-term activation and expansion for T cells enriched stem cell memory-like T cells, enhanced glycolytic metabolism, and improved the persistence of CAR-T cells[41,42]. Overall, these strategies are expected to further improve the clinical activity of CLDN18.2-targeting CAR-T cell therapies. Despite encouraging preclinical and early clinical data, evidence for CAR-T therapy in CLDN18.2-positive PDAC is still limited. Rigorous phase I/II trials with defined endpoints for stromal penetration and T-cell persistence are urgently needed.

PDAC-specific milestone summary

CT041 filed a marketing application; its pancreatic cancer indication is under review, with phase Ib adjuvant and phase Ib/II advanced trials ongoing. MC-008 has been granted orphan drug designation by the United States FDA for the treatment of pancreatic cancer. Currently, PDAC-specific evidence remains limited to small-scale investigations, and rigorous trials focusing on matrix penetration and T-cell persistence are urgently needed to further validate these therapies.

BsAbs

To analyze the heterogeneity of expression subtypes in PDAC and explore co-expression, multiplex immunofluorescence to detect six subtype markers, including CLDN18.2, was performed on 289 primary PDAC samples, 37 metastatic lesions, and 77 patient-derived organoids[85]. The results revealed significant subtype heterogeneity in PDAC and that co-expressing cells may represent a transitional state. These findings suggest that targeted therapy against a single subtype may have limited efficacy, and thus, PDAC may require combination therapy or the development of strategies targeting the transitional state. Representative CLDN18.2-targeting BsAb drugs include AZD5863[86], givastomig[87], gresonitamab (AMG 910)[88], IBI389[89], PM1032, PT886, Q-1802, and QLS31905[90]. Compared with anti-CLDN18.2 mAbs and ADCs, anti-CLDN18.2 BsAbs are in a relatively early stage of clinical development[86]. For example, givastomig, IBI389, and Q-1802 are being tested in phase I clinical trials; AZD5863, PM1032, and PT886 are being tested in phase I/II clinical trials; and only QLS31905 is being tested in a phase III trial. QLS31905 is not only the fastest-advancing BsAb drug but also the first CLDN18.2/CD3 BsAb to advance to a PDAC-dedicated registrational phase III trial. The latest data from a phase I study of the use of QLS31905 to treat solid tumors were presented at the 2025 ASCO Annual Meeting[91]. Among 33 CLDN18.2-positive solid tumor patients who received treatment, the ORR was 18.18% (6/33, all 6 cases achieved PR), and the DCR was 87.88% (29/33). The mPFS and median OS were 4.21 months and 9.53 months, respectively. For a specific subset of patients, the ORR was 25.00% (3/12), the DCR was 91.67% (11/12), the mPFS was 3.94 months, and the median OS had not been reached yet. In July 2025, a randomized, double-blind, multicenter phase III trial (NCT07079228) was initiated to evaluate QLS31905 in combination with nab-paclitaxel plus gemcitabine vs placebo plus chemotherapy as first-line treatment for CLDN18.2-positive advanced pancreatic cancer, with a planned enrollment of 602 patients. The first patient was dosed on September 8, 2025. At the 2026 ASCO Gastrointestinal Cancers Symposium[92], results from the TWINPEAK study of spevatamig (PT886), a CLDN18.2/CD47 BsAb, in combination with gemcitabine/nab-paclitaxel for first-line metastatic PDAC were reported in 42 patients. Among the 2 mg/kg weekly cohort (n = 15), treatment resulted in an ORR of 40%, DCR of 93%, mPFS of 7.3 months, and median OS of 13.2 months. Updated data presented at the 2026 ASCO Annual Meeting included an ORR of 52.4%, a DCR of 90.5%, and a median OS of 14.7 months. Of note, 85% of screened patients met the threshold of CLDN18.2 expression ≥ 10% (≥ 2+ staining). Spevatamig has received United States FDA orphan drug designation for pancreatic cancer, and these data strongly support the further development of a randomized pivotal phase III trial.

Among the CLDN18.2-targeting BsAbs, AZD5863, gresonitamab, IBI389, and QLS31905 can simultaneously bind to CLDN18.2 on tumor cells and CD3 on T cells[86]. Their mechanism of action involves activating CD3 signaling in T cells to directly recruit T cells to the vicinity of tumor cells, where they can form lytic synapses and release granzyme and perforin to specifically attack tumor cells[93,94]. These agents are considered bispecific T-cell engagers (BiTEs), a subset of BsAbs. Compared with traditional Abs, BiTEs have a smaller molecular weight, enabling more efficient penetration of tumor tissues[58]. Their Fc domain helps maintain stability, simplify the purification process, and extend their half-life[95]. A study of IBI389 in patients with refractory advanced PDAC enrolled 120 patients who met the criterion that ≥ 10% of tumor cells showed IHC 2+/3+ CLDN18.2 expression. A clinical response was observed in 27 out of 91 CLDN18.2-positive patients, with an ORR of 29.6%, a DCR of 70.4%, and a mPFS not yet reached at the time of preliminary result reporting due to limited follow-up duration[89].

Although CLDN18.2-targeting BiTEs can inhibit tumor growth in the early stage, their efficacy significantly diminishes for advanced cancers[95]. However, the Fc fragment of BiTEs interacts with CD64+ CAFs, and through the activation of the spleen tyrosine kinase-Vav guanine nucleotide exchange factor 2 (VAV2)-Ras homolog family member A-Rho-associated protein kinase-myosin light chain 2-myocardin-related transcription factor A-SMA/collagen I pathway, fibrosis is enhanced, which restricts T-cell infiltration[95]. Conversely, the activity of the cyclic guanosine monophosphate-AMP synthase/stimulator of interferon genes (STING) pathway in tumor-infiltrating lymphocytes decreases, leading to a reduction in T cell factor 1 + PD-1+ stem cell-like CD8+ T cells, which impairs the long-term effect of BiTEs[48]. Further studies[95] have shown that the long-acting β2-adrenergic receptor agonist vilanterol can attenuate BiTE-induced fibrosis by inhibiting the phosphorylation of VAV2 (Y172) in CD64+ CAFs. In contrast, STING agonists can promote the proliferation of stem cell-like CD8+ T cells. When both are used in combination with CLDN18.2 BiTEs, they exert a synergistic effect, inhibiting the activation of CD64+ CAFs, increasing the number of stem cell-like CD8+ T cells, and achieving sustained anti-tumor activity, thereby providing a potential new strategy for the treatment of PDAC. In addition, a study innovatively constructed an oncolytic herpes simplex virus type 1 BiTE (OV-BiTE) carrying a CLDN18.2/CD3 BiTE. This OV-BiTE was found to induce immunogenic tumor cell death, activate T cells, and promote the secretion of cytokines such as interferon gamma (IFN-γ) and IL-2. For in vivo experiments, two models were established: The KPC [LSL-KRAS^G12D/+; LSL-Trp53^R172H/+; Pdx-1-cyclization recombinase (Cre)] model, which is highly similar to the pathological characteristics of human PDAC and driven by KRAS mutation, p53 mutation, and Cre recombinase expression; and the Pan02 model, which originates from the malignant transformation of mouse pancreatic acinar cells, with relatively rapid tumor growth and slightly higher immunogenicity than the KPC model. In these two syngeneic pancreatic cancer models, OV-BiTE significantly inhibited tumor growth compared with the control group. Moreover, the mPFS of the KPC model with OV-BiTE treatment was 42 days, which was longer than the PFS of 25.5 days in the control group, indicating that OV-BiTE can prolong the survival of mice without obvious hepatorenal toxicity. The oncolytic effect of oncolytic viruses provides BiTEs with more recognizable tumor antigens and space for T-cell infiltration, while BiTEs amplify the adaptive immune response triggered by the virus, forming a triple effect of “direct oncolysis + immune activation + targeted killing”. These results indicate that OV-BiTE can effectively overcome the immunosuppressive barrier of the pancreatic cancer TME and provide a preliminary basis for clinical research[96]. Most drugs are still in phase I/II clinical trials, and specific data regarding their efficacy for PDAC are lacking. IBI389 achieved an ORR of 29.6% in PDAC, but mPFS was not reached at reporting, raising questions about follow-up duration and durability. Despite mechanistic innovations, the clinical benefit of BsAbs in PDAC requires further validation in larger, PDAC-specific trials.

PDAC-specific milestone summary

QLS31905 has advanced to a PDAC-dedicated phase III registrational trial. For PT886, a PDAC-specific phase II study has been completed and supports pivotal development. The phase I/II program for givastomig has incorporated PDAC expansion cohorts. Other CLDN18.2-targeting BsAbs remain in early-phase (I/II) trials without PDAC-specific registrational data reported. Overall, these agents have progressed from signal-seeking to pivotal-validation stages in PDAC, though final clinical benefit remains to be determined by larger RCTs.

Others

In addition to the four targeted therapy approaches described above, several studies have explored new therapeutic methods. All of these approaches are summarized in Table 2. The trispecific antibody DR30318 was designed to simultaneously bind to CLDN18.2, human serum albumin (HSA), and CD3 on the T-cell surface, and in vitro experiments have shown that DR30318 exhibits high affinity for CLDN18.2, CD3, and HSA (half maximal effective concentration values of 14.4 nM, 3.14 nM, and 0.44 nM, respectively). Further experiments showed it can induce potent T cell-dependent cellular cytotoxicity (TDCC), with particularly significant effects on gastric cancer and pancreatic cancer cells with low/medium CLDN18.2 expression, and can also promote T-cell proliferation. In in vivo experiments, DR30318 exhibited significant tumor-suppressive effects in both gastric cancer (NUGC4/CLDN18.2-overexpressing) and pancreatic cancer (BxPC3/hCLDN18.2) xenograft models. Complete tumor regression was achieved with some doses, with no impact on the body weight of mice. Owing to its trispecific design, which extends its half-life, enhances potent TDCC activity, and reduces the risk of cytokine release, DR30318 holds promise as a potentially effective therapeutic agent for CLDN18.2-positive gastric and pancreatic cancers[97].

Table 2 Summary of research progress on claudin 18.2-targeting drugs in pancreatic cancer.

Drug class	Drug name	Company	Indications	Clinical testing phase	Registration number
mAb	Zolbetuximab	Astellas Pharma/Baxter Oncology GmbH	CLDN18.2-positive metastatic pancreatic adenocarcinoma/pancreatic adenocarcinoma	2	CTR20220914
		Astellas Pharma Global Development, Inc.	Pancreatic adenocarcinoma metastatic/CLDN18.2-positive pancreatic adenocarcinoma	1	NCT06396091
	FG-M108	FutureGen Biopharmaceutical Co., Ltd.	Pancreatic cancer	3	NCT07383922/CTR20260995
	Osemitamab	Suzhou Transcenta Therapeutics Co., Ltd.	PDAC	2	NCT04495296
		Bristol Myers Squibb Co.	Pancreatic cancer	2	NCT04396821
	ASKB589	Jiangsu Aosaikang Biomedicine Co., Ltd.	Pancreatic cancer	2	CTR20202121
	AB-011	CARsgen Therapeutics Co., Ltd.	Pancreatic adenocarcinoma	1	NCT04400383
	FL-301	Flame Biosciences LLC	Pancreatic cancer	2	NCT05181865
ADC	CMG901 (AZD0901)	Astra Zeneca PLC	PDAC	2	NCT06219941
		Keymed Biomedical Technology (Chengdu) Co. Ltd.	Pancreatic cancer	1	NCT04805307
	ATG-022	Antengene (Hangzhou) Biologics Co. Ltd.	Pancreatic cancer	Preclinical	FDA
	BL-M05D1	Systimmune, Inc.	Advanced PDAC	1	NCT07021066
	SHR-A1904	Shanghai HengruiPharmaceutical Co., Ltd.	Advanced pancreatic adenocarcinoma	1	CTR20211290
	IBI343	InnoventBiologics (Suzhou) Co., Ltd.	PDAC	1	NCT5458219
			PDAC	1	NCT05458219
			Pancreatic cancer	3	NCT07066098
	E0-3021	CSPC Pharmaceutical Group Ltd./Elevation Oncology, Inc.	Pancreatic cancer	1	The details of the drug have not been disclosed
ADC	XNW27011	Evopoint Biosciences Co., Ltd.	Metastatic pancreatic adenocarcinoma	2	NCT06792435
	JS107	Shanghai Junshi Biosciences Co., Ltd.	Pancreatic cancer	1	CTR20223134/NCT05657418
	LM-302	LaNova Medicines (Zhejiang) Co., Ltd.	Pancreatic cancer	3	The details of the drug have not been disclosed
	SOT-102	SOTIO Biotech as	Advanced pancreatic adenocarcinoma/pancreatic adenocarcinoma	2	NCT05525286
		Excoso as	Pancreatic cancer	2	The details of the drug have not been disclosed
CAR-T	IMC-008	Suzhou Immunofoco Biotechnology Co., Ltd.	Pancreatic cancer	The details of the drug have not been disclosed	FDA
	SENL-215	Hebei Senlangbio Biotechnology Co., Ltd.	Pancreatic cancer	Preclinical	The details of the drug have not been disclosed
	DCTY-1502	Beijing Dingcheng Peptide Source Biotech Co., Ltd.	Pancreatic cancer	Preclinical	The details of the drug have not been disclosed
BsAb	SNG2002	Beijing Shenogen Pharma Group Ltd.	Pancreatic cancer	Preclinical	The details of the drug have not been disclosed
	QLS31905	Qilu Pharmaceutical Co., Ltd.	Pancreatic cancer	3	NCT07079228

mAb: Monoclonal antibodies; ADC: Antibody drug conjugates; BsAb: Bispecific antibodies; CAR-T: Chimeric antigen receptor T cell; CLDN18.2: Claudin 18.2; PDAC: Pancreatic ductal adenocarcinoma; FDA: United States Food and Drug Administration; CTR: Clinical trial registration; NCT: National clinical trial.

A genetically engineered membrane-based nano-connector (NanoBE) was constructed by anchoring anti-CD40 single-chain variable fragment (scFv) targeting macrophages and anti-CLDN18.2 scFv on the cell membrane, followed by encapsulation in a poly (lactic-co-glycolic acid) nano-core[98]. It was shown to significantly enhance the recognition and phagocytosis of CLDN18.2-positive pancreatic cancer cells by macrophages, to increase the expression of CD80/CD86, and to promote macrophage activation and antigen presentation. As a result, IFN-γ secretion was increased by 4-fold, and tumor growth was significantly inhibited in pancreatic cancer models, with survival extended to more than 100 days. This NanoBE provides a new immunotherapeutic strategy for “cold” pancreatic cancer, which has low immunogenicity and weak anti-tumor immune response.

Another reported strategy involves a chimeric hepatitis B virus core antigen virus-like particle (VLP) vaccine[99]. In mice and rabbits, this vaccine can induce autoantibodies that bind specifically to CLDN18.2 on the surface of tumor cells, with no cross-reactivity with CLDN18.1, which differs by only one amino acid. It was shown to efficiently kill CLDN18.2-positive cells in vitro by inducing ADCC and CDC. Additionally, in mouse models, it partially inhibited the metastatic growth of CLDN18.2-positive tumor cells to the lungs. This research provides proof-of-concept for the application of VLP vaccines in cancer immunotherapy. Despite lacking crossreactivity with CLDN18.1, this approach carries a theoretical risk of autoimmunity against normal gastric mucosa where CLDN18.2 is physiologically expressed, and long-term safety studies are lacking. All three novel approaches remain at preclinical or very early stages of development, and none have been specifically validated in PDAC models that recapitulate the dense desmoplastic stroma and immunosuppressive microenvironment. Thus, these strategies require further validation in more clinically relevant PDAC models and rigorous safety assessments before clinical translation.

CONCLUSION

The current understanding of the role of CLDN18.2 in PDAC is far from mature. Most data are exploratory, and caution is needed when interpreting early-phase results. PDAC has a high degree of aggressiveness among all malignancies, and its clinical treatment is challenging due to significant limitations in surgical success, a low cure rate, and its traditional target, KRAS, having long been regarded as an “undruggable target”. As a novel target with high expression, targetability, and low off-target effects, CLDN18.2 has provided a new therapeutic strategy for PDAC and filled a major gap in the field of targeted therapy. CLDN18.2 can not only directly kill tumor cells but also participates in the occurrence and development of tumors through its barrier function and regulation of the tumor immune microenvironment. Although most studies on CLDN18.2 have focused on its roles in G/GEJ cancer, clinical studies on the treatment of CLDN18.2-positive pancreatic cancer patients have been gradually reported in recent years. The mAb zolbetuximab, targeting CLDN18.2, is currently the first and only drug approved for clinical use. Research to develop other therapeutic methods is still mostly in the phase I/II clinical trial stages, but their findings are verifying the efficacy of relevant drugs for PDAC treatment. The great potential of CLDN18.2 as a diagnostic and therapeutic target makes it an important research direction in the field of PDAC precision therapy. However, several PDAC-specific challenges must be addressed before these strategies can be widely adopted. First, the dense desmoplastic stroma, composed of CAFs and extracellular matrix, physically restricts the penetration of antibodies, ADCs, and CAR-T cells. Second, the highly immunosuppressive microenvironment, marked by abundant Tregs, MDSCs, and inhibitory cytokines (TGF-β, IL-10), dampens the activity of T cell-engaging therapies. Third, antigen heterogeneity and escape, including clonal loss of CLDN18.2 expression or switching to CLDN18.1, may drive acquired resistance. As discussed in this review, several strategies have been explored to overcome these barriers, including stromal modulators or nanoparticle-based delivery systems to enhance drug penetration, as well as dual-targeting CAR-T designs and BiTEs to improve T-cell function and reduce exhaustion. Co-targeting immune checkpoints or myeloid suppressors represents a rational but largely unexplored strategy that could synergize with CLDN18.2-directed agents, warranting future preclinical and clinical investigation. Future research should prioritize PDAC-specific preclinical models recapitulating these features and incorporate biomarker-driven trials with mandatory stromal and immune profiling. It must be reiterated that the clinical evidence for CLDN18.2-targeted therapies in PDAC is still at an early stage. Most data are exploratory, and caution is needed when interpreting early-phase results.