Executive Summary
Claudin 18.2 (CLDN18.2) has evolved from a niche research target into one of the most clinically validated biomarkers in gastrointestinal oncology. Over the period 2020–2026, the field has witnessed the regulatory approval of a first-in-class monoclonal antibody (mAb), the emergence of multiple late-stage antibody–drug conjugates (ADCs), the first-ever solid-tumor pivotal CAR-T cell therapy trial publication, and a rapidly expanding bispecific antibody ecosystem. The October 2024 U.S. Food and Drug Administration (FDA) approval of zolbetuximab-clzb (Vyloy; Astellas Pharma) — followed within weeks by European Medicines Agency (EMA) authorization — established CLDN18.2 as a commercialized therapeutic axis in gastric and gastroesophageal junction (GEJ) adenocarcinoma 1415. Parallel development of ADCs, CAR-T therapies, and bispecific antibodies now defines the competitive landscape for 2026 and beyond 16.
Mechanistic and Biomarker Background
CLDN18.2 is a tight-junction protein expressed selectively in differentiated gastric mucosal epithelial cells, with negligible expression in undifferentiated gastric stem cells and other healthy tissues 3. During malignant transformation, disrupted cell polarity exposes CLDN18.2 epitopes previously inaccessible in the basolateral compartment, creating a targetable tumor-associated antigen with a favorable on-target/off-tumor selectivity profile 37.
Tumor expression landscape. CLDN18.2 is expressed in gastric cancer (GC), GEJ adenocarcinoma, pancreatic ductal adenocarcinoma (PDAC), esophageal cancer, ovarian adenocarcinoma, and selected lung cancers 3. Reported positivity rates vary substantially depending on the immunohistochemistry (IHC) threshold applied:
- Using the stringent threshold of moderate-to-high membranous staining in ≥75% of tumor cells (as in SPOTLIGHT and GLOW trials), approximately 24–38% of GC/GEJ adenocarcinomas are positive 6.
- Using the same threshold in PDAC, approximately 30.4–32.5% of cases qualify, suggesting that roughly one-third of pancreatic cancer patients may be eligible for CLDN18.2-directed therapy 6.
- In biliary tract cancer (BTC), positivity at the ≥75% threshold is observed in only 5.5% of cases overall, with higher rates in gallbladder cancer (14.6%) and extrahepatic cholangiocarcinoma (8.6%) compared to intrahepatic subtypes (2.0%) 6.
Biomarker standardization challenges. Intratumoral heterogeneity in CLDN18.2 expression affects 33–40% of GC cases, and primary-to-metastasis discordance occurs in 20–25% of patients — mirroring patterns observed with HER2 6. The FDA-approved VENTANA CLDN18 (43-14A) RxDx Assay has become the primary standardized diagnostic tool, but discrepancies among laboratories and the absence of a universally validated positivity cutoff remain unresolved 614. Notably, the FAST trial used a lower threshold of ≥40% of cells with 2+/3+ staining, whereas SPOTLIGHT and GLOW used ≥75%; subgroup analyses consistently suggest that higher CLDN18.2 expression thresholds enrich therapeutic benefit 46.
Benchmark Monoclonal Antibody: Zolbetuximab
Zolbetuximab is a chimeric IgG1 mAb that binds the extracellular CLDN18.2 domain and mediates tumor cell death through antibody-dependent cellular cytotoxicity (ADCC) and complement-dependent cytotoxicity (CDC) 4. Its clinical validation across pivotal phase III trials established the efficacy and safety benchmark against which all subsequent modalities are positioned 14:
- SPOTLIGHT trial (NCT03504397): 565 patients randomized to zolbetuximab + mFOLFOX6 versus placebo + mFOLFOX6. Median progression-free survival (PFS): 10.6 vs. 8.7 months (hazard ratio [HR] 0.751; p=0.0066); median overall survival (OS): 18.2 vs. 15.5 months (HR 0.750; p=0.0053) 14.
- GLOW trial (NCT03653507): 507 patients randomized to zolbetuximab + CAPOX versus placebo + CAPOX. Median PFS: 8.2 vs. 6.8 months (HR 0.687; p=0.0007); median OS: 14.4 vs. 12.2 months (HR 0.771; p=0.0118) 14.
The most common adverse events were nausea (82%), vomiting (67%), and decreased appetite (47%), consistent with on-target gastrointestinal (GI) toxicity from CLDN18.2 expression in normal gastric mucosa 414. Zolbetuximab received both priority review and orphan drug designation; concurrent FDA approval of the companion diagnostic established the required patient-selection framework 14.
Antibody–Drug Conjugates
ADCs represent the most crowded late-stage investigational modality in the CLDN18.2 field, combining CLDN18.2-directed antibodies with cytotoxic payloads to enable targeted intracellular drug delivery 1.
LM-302 (tecotabart vedotin; LaNova Medicines/Bristol Myers Squibb) is the most clinically advanced CLDN18.2 ADC by enrollment milestone. As of February 2026, LM-302 completed patient enrollment in a phase III registrational trial (LM302-03-101) for third-line or later CLDN18.2-positive gastric/GEJ adenocarcinoma — the first CLDN18.2-directed ADC to reach this milestone globally 17. At ASCO 2025, combination data with toripalimab (anti-PD-1) in gastric cancer demonstrated an objective response rate (ORR) of 65.9% and disease control rate (DCR) of 85.4% among 41 evaluable patients; in the subgroup with CLDN18.2 ≥25%, ORR reached 71.9% and DCR 96.9% 17. Phase III development in first-line gastric/GEJ cancer combined with a PD-1 inhibitor is planned 17.
IBI-343 (Innovent Biopharmaceuticals) is a fully humanized, Fc-silent anti-CLDN18.2 ADC conjugated to exatecan via site-specific glycol conjugation with a cleavable linker and a drug-to-antibody ratio of 4. The Fc-silent design is intended to minimize ADCC/CDC and potentially reduce on-target GI toxicity 34. In the published phase I study (NCT05458219; 127 patients: 19 escalation, 108 expansion), the recommended phase 2 dose (RP2D) was 6 mg/kg every 3 weeks. At RP2D in CLDN18.2-high GC/GEJ patients (defined as 2+/3+ staining in ≥75% of cells), confirmed ORR was 29.0–32.3%, DCR 90.3%, median duration of response (DoR) 5.6 months, and median PFS 5.5 months 34. Hematologic toxicity was the dominant safety signal: grade ≥3 treatment-related adverse events (TRAEs) occurred in 52.6%; notably, no interstitial lung disease (ILD) was observed 34. IBI-343 has received FDA Fast Track Designation for advanced PDAC 34.
AZD-0901 (CMG901; AstraZeneca/Chengdu KeyMed/Lepu Biopharma) is a phase III CLDN18.2-targeted ADC with activity across gastric/GEJ cancer, PDAC, and BTC. Secondary summaries from the available dataset report ORR of 75.0% and DCR of 100% in CLDN18.2-positive gastric/GEJ cancer, with PFS ranging from 3.3–14.5 months 34. A randomized later-line study (NCT06346392) has been registered, signaling progression toward registrational evidence generation 34. However, a peer-reviewed primary efficacy publication was not available in the retrieved materials, and numerical data should be interpreted with caution.
EO-3021 (SYSA-1801; CSPC Pharmaceutical/Elevation Oncology) is another mid-stage ADC evaluated across the US, China, and Japan; detailed efficacy data were not available in the retrieved materials 1.
LM-302 (phase I/II single-arm ADC data): An earlier dataset from 135 enrolled patients (36 evaluable at RP2D) demonstrated ORR 30.6%, DCR 75.0%, and median PFS 7.16 months in ≥2nd-line gastric/GEJ cancer at a recommended phase II dose of 1.8 mg/kg every 2 weeks; hematologic toxicity was the primary safety concern, with grade ≥3 neutropenia in 22.2% 2.
CAR-T Cell Therapies
CLDN18.2-directed CAR-T therapies represent a high-risk/high-potential modality that has achieved proof-of-concept in solid tumors — a setting historically resistant to cell therapy approaches 4.
Satricabtagene autoleucel (satri-cel; CT041; CARsgen/Moderna) is the most advanced CAR-T asset in the CLDN18.2 field. Phase I interim results (NCT03874897; 37 patients with CLDN18.2-positive digestive system cancers) reported: overall ORR 48.6%, DCR 73.0%; in gastric cancer specifically, ORR 57.1%, DCR 75.0%, 6-month OS rate 81.2% 34. In the optimal dose/pretreated gastric subgroup (≥2 prior lines, 2.5 × 10^8 cells), ORR reached 61.1%, DCR 83.3%, and median PFS 5.6 months 34. Safety was characterized by universal grade ≥3 hematologic toxicity attributable to the preconditioning lymphodepletion regimen (fludarabine, cyclophosphamide, nab-paclitaxel); grade 1–2 cytokine release syndrome (CRS) occurred in 94.6% of patients, with no grade ≥3 CRS, no grade ≥3 neurotoxicity (immune effector cell-associated neurotoxicity syndrome, ICANS), and no treatment-related deaths 434. Pivotal phase II data (CT041-ST-01) were published in The Lancet in 2025, representing the first CLDN18.2-directed CAR-T therapy to achieve this milestone; specific numerical endpoints from the final publication were not reproduced in full in the retrieved materials 16.
Two documented cases in PDAC demonstrated CT041-mediated partial responses (PR) and one complete response (CR) in CLDN18.2-positive pancreatic cancer, with CAR copy number expansion and durable disease control 5. Preclinical strategies to enhance solid-tumor CAR-T efficacy include sequential FAP-targeted CAR-T infusion to deplete immunosuppressive cancer-associated fibroblasts (CAFs) and suppress myeloid-derived suppressor cell (MDSC) recruitment before CLDN18.2-directed CAR-T administration 10, and transgenic IL-15 armoring (H9 CAR-IL15) to improve persistence, reduce exhaustion, and enhance in vivo expansion — though antigen-negative escape emerged as an immune evasion mechanism 11.
Multiple additional autologous and allogeneic CAR-T programs (LB-1908, IM-92, IMC-002, GC-506, KD-496) remain in phase I development, predominantly in China 1.
Bispecific Antibodies and T-Cell Engagers
The bispecific antibody segment in the CLDN18.2 landscape is mechanistically diverse, pairing CLDN18.2 with costimulatory, immune checkpoint, or phagocytic targets.
Givastomig (ABL-111/TJ033721; Astellas/I-Mab) is a CLDN18.2 × 4-1BB bispecific antibody. At ESMO GI 2025 (Barcelona, July 2–5, 2025), phase Ib combination data with nivolumab and mFOLFOX6 in CLDN18.2-positive (≥1+ intensity in ≥1% of tumor cells) gastric/esophageal/GEJ adenocarcinoma demonstrated ORR of 71% and DCR of 100% across dose cohorts; 6-month PFS rate was 72.9% (81.5% at ≥8 mg/kg) 18. No dose-limiting toxicities (DLTs) were observed. TRAEs included neutropenia (71%), peripheral neuropathy (59%), nausea (53%), infusion-related reactions (IRR, 41%), and vomiting (35%) 18. The broader CLDN18.2 positivity cutoff (≥1% vs. ≥75%) may substantially expand the eligible patient population but complicates cross-trial efficacy comparisons.
Spevatamig (PT-886; Phanes Therapeutics) is an anti-CLDN18.2/CD47 bispecific antibody that simultaneously blocks CLDN18.2 and the phagocytic "don't-eat-me" signal mediated by CD47. In the TWINPEAK study (NCT05482893), first-line mPDAC patients treated with spevatamig 2 mg/kg weekly plus gemcitabine/nab-paclitaxel (GnP; n=15) achieved ORR 40%, DCR 93%, median PFS 7.3 months, and median OS 13.2 months 19. Critically, no spevatamig-attributed grade ≥3 hematologic or GI toxicity and no CRS were observed, suggesting a highly differentiated tolerability profile 19. Responses were observed across patients with CLDN18.2 ≥10% (≥2+ staining), and 85% of screened patients met this eligibility threshold 19. Spevatamig has received both orphan drug designation (2022) and FDA Fast Track Designation (2024) for pancreatic cancer 19.
ASP-2138 (Astellas/Xencor) and AMG-910 (Amgen/BeiGene) are additional bispecific programs in phase I development with multinational clinical footprints spanning the US, Japan, France, Germany, and China 1. Q-1802 (QureBio) is a mid-stage CLDN18.2 × PD-L1 bispecific in phase II development in China and the US 1.
Preclinically, the CD3/CLDN18.2 XFab bispecific ZWB67 demonstrated high T-cell–redirected cytotoxicity and in vivo efficacy in CLDN18.2-expressing xenograft models with low-affinity anti-CD3 design to mitigate systemic T-cell activation 9.
Competitive Landscape Table
| Asset | Company | Modality | Targets | Indication(s) | Phase / Status | Key Regulatory Milestones | Geography |
|---|---|---|---|---|---|---|---|
| Zolbetuximab (Vyloy) | Astellas Pharma | mAb | CLDN18.2 | GC/GEJ (1L, HER2-) | FDA/EMA Approved (Oct/Sep 2024) | Priority Review; Orphan Drug; CDx approval | Global |
| LM-302 (tecotabart vedotin) | LaNova / BMS | ADC | CLDN18.2 | GC/GEJ (3L+); 1L planned | Phase III (enrollment complete Feb 2026) | Breakthrough Therapy (CDE China); Orphan Drug (FDA) | China, USA |
| AZD-0901 (CMG901) | AstraZeneca / KeyMed | ADC | CLDN18.2 | GC/GEJ, PDAC, BTC | Phase II/III | Pan-tumor expansion; randomized study registered | USA, UK, Japan, China, Spain |
| IBI-343 | Innovent | ADC | CLDN18.2 | GC/GEJ, PDAC, BTC | Phase III | FDA Fast Track (PDAC) | China, Japan, USA |
| EO-3021 (SYSA-1801) | CSPC / Elevation Oncology | ADC | CLDN18.2 | GC/GEJ | Phase II | — | USA, China, Japan |
| Satri-cel (CT041) | CARsgen / Moderna | CAR-T | CLDN18.2 | GC/GEJ, PDAC | Phase II (pivotal; Lancet 2025) | — | China, USA |
| Givastomig | Astellas / I-Mab | Bispecific (CLDN18.2 × 4-1BB) | CLDN18.2, 4-1BB | GC/GEJ/Esophageal | Phase Ib | — | China, USA |
| Spevatamig (PT-886) | Phanes Therapeutics | Bispecific (CLDN18.2 × CD47) | CLDN18.2, CD47 | PDAC (1L) | Phase Ib/II | Orphan Drug (2022); Fast Track (2024) | USA, China |
| ASP-2138 | Astellas / Xencor | Bispecific | CLDN18.2 | Solid tumors | Phase I | — | USA, Japan, France, Italy, China |
| AMG-910 | Amgen / BeiGene | Bispecific | CLDN18.2 | Solid tumors | Phase I | — | USA, EU, Japan, China |
| Q-1802 | QureBio | Bispecific (CLDN18.2 × PD-L1) | CLDN18.2, PD-L1 | GC/GEJ | Phase II | — | China, USA |
| ASKB-589 | Jiangsu Aosaikang | mAb | CLDN18.2 | GC/GEJ | Phase III | — | China |
| FG-M108 | FutureGen Biopharm | mAb | CLDN18.2 | GC/GEJ, Pancreatic | Phase III/I | — | China |
| Osemitamab (TST001) | Transcenta | mAb | CLDN18.2 | GC/GEJ | Phase II | — | China, USA |
Clinical Data Summary Table
| Asset | Setting / Line | Evaluable n | ORR | DCR | Median PFS | Median OS | Key Safety Signal |
|---|---|---|---|---|---|---|---|
| Zolbetuximab + mFOLFOX6 (SPOTLIGHT) | 1L GC/GEJ | 565 (randomized) | 60.7% | — | 10.6 mo | 18.2 mo | Nausea 82%, vomiting 67%, neutropenia |
| Zolbetuximab + CAPOX (GLOW) | 1L GC/GEJ | 507 (randomized) | — | — | 8.2 mo | 14.4 mo | Nausea, vomiting, diarrhea |
| LM-302 monotherapy (Phase I/II) | ≥2L GC/GEJ | 36 (evaluable) | 30.6% | 75.0% | 7.16 mo | 6-mo OS 95% | Neutropenia Gr≥3 22.2%; WBC decrease 17.8% |
| LM-302 + toripalimab (ASCO 2025) | GC/GEJ | 41 | 65.9% | 85.4% | NR | NR | Manageable; specific Gr≥3 not detailed |
| IBI-343 6 mg/kg Q3W | CLDN18.2-high GC/GEJ | ~31 | 29.0–32.3% | 90.3% | 5.5 mo | NR | Hematologic Gr≥3 52.6%; no ILD |
| AZD-0901 | CLDN18.2+ GC/GEJ | Not detailed | ~75.0%* | ~100%* | 3.3–14.5 mo* | 8.5 mo–NR* | Favorable (secondary source) |
| CT041 (Phase I; all tumors) | ≥2L digestive cancers | 37 | 48.6% | 73.0% | NR | NR | CRS Gr1-2 94.6%; universal hematologic (conditioning) |
| CT041 (Phase I; GC/GEJ subset) | ≥2L GC/GEJ | 28 | 57.1% | 75.0% | 5.6 mo | 6-mo OS 81.2% | No Gr≥3 CRS, no ICANS |
| Givastomig + nivo + mFOLFOX6 | 1L GC/GEJ/Esophageal | 17 (across doses) | 71% | 100% | 6-mo PFS 72.9–81.5% | NR | Neutropenia 71%, peripheral neuropathy 59% |
| Spevatamig 2 mg/kg QW + GnP | 1L mPDAC | 15 | 40% | 93% | 7.3 mo | 13.2 mo | No spevatamig-attributed Gr≥3 hematologic/GI toxicity; no CRS |
| FG-M108 + CAPOX (medium/high CLDN18.2) | 1L GC/GEJ | ~36 | 77.8% | 97.2% | 9.6 mo | NR | Hematologic, GI (chemo-related); no Gr4/5 mAb-related |
| FG-M108 + GnP (mod/high CLDN18.2) | 1L PDAC | 32 | 53.1% | 100% | 9.9 mo | 17.4 mo (immature) | Gr≥3 TEAEs 46.9%; no Gr4/5 mAb-related AEs |
*AZD-0901 data derived from secondary summaries; interpret with caution 34.
Outlook and Key Takeaways
The CLDN18.2 field is no longer a single-asset story. Zolbetuximab's dual regulatory approval in 2024 validated the target but also revealed the ceiling of naked antibody monotherapy: median OS gains of approximately 2–3 months over chemotherapy alone underscore a persistent unmet need 144. The competitive landscape is now defined by payload-enabled modalities (ADCs), engineered cell therapies (CAR-T), and multi-target bispecifics that seek to surpass this efficacy threshold 16.
ADCs are the most crowded and rapidly advancing class. LM-302's phase III enrollment completion in early 2026 positions it as the first likely second regulatory approval in this target class 17. IBI-343's differentiated Fc-silent/exatecan design may offer a GI-toxicity advantage over ADCC/CDC-active formats 34. AZD-0901's pan-tumor expansion into PDAC and BTC further widens the potential addressable population 34.
CAR-T therapy has demonstrated proof-of-concept in solid tumors, but key barriers remain: mandatory lymphodepletion, manufacturing logistics, vein-to-vein time, CRS management, antigen escape (CLDN18.2 downregulation under immune pressure), and lack of scalable redosing strategies 1134. The publication of satri-cel's phase II data in The Lancet represents a landmark advance; broader adoption will depend on long-term durability and commercialization feasibility 16.
Bispecifics offer mechanistic novelty. Spevatamig's CD47 blockade strategy — with no spevatamig-attributed grade ≥3 hematologic toxicity and no CRS in the pancreatic cancer combination trial — represents a potentially differentiated tolerability profile 19. Givastomig's 4-1BB costimulation showed striking early ORR, though small sample size and low CLDN18.2 entry threshold limit cross-trial interpretation 18.
Biomarker harmonization is the most critical unresolved infrastructure challenge. The divergence between CLDN18.2 positivity thresholds (from ≥1% at any intensity in givastomig trials to ≥75% at 2+/3+ in SPOTLIGHT/GLOW) creates fundamental cross-trial incomparability and inconsistent real-world patient selection 618. Until ROC-optimized or analytically validated universal cutoffs are established, clinicians and developers must interpret all efficacy comparisons across CLDN18.2-targeted assets with caution 6.
Geographic concentration. China is the dominant innovation hub across all modalities and development stages, with the largest number of phase I–III programs and the broadest pipeline breadth 1. The US, Japan, and Europe serve primarily as multinational trial expansion and commercialization geographies 1.
For clinicians: CLDN18.2 IHC testing with the VENTANA CDx assay should be considered standard workup for all patients with locally advanced or metastatic GC/GEJ adenocarcinoma. Enrollment in clinical trials evaluating next-generation ADCs and bispecifics — particularly for later-line or PDAC settings — represents the most meaningful pathway to improve on zolbetuximab-established benchmarks. For PDAC specifically, spevatamig's promising combination data and the feasibility of CT041-mediated responses suggest that CLDN18.2 screening in pancreatic cancer is increasingly warranted 519.