This example study evaluates the safety, feasibility, cellular kinetics, and preliminary anti-tumor activity of EBNK-1822H2, an illustrative allogeneic cord blood-derived dual-target CAR-NK cell product directed against CLDN18.2 and HER2, in adults with relapsed/refractory or metastatic esophageal adenocarcinoma after standard therapy. The study uses dose escalation followed by biomarker-defined expansion and prospectively records EGFR expression as an exploratory biomarker of antigen escape.

Background and target-selection logic. Esophageal adenocarcinoma shares biomarker biology with GEJ adenocarcinoma. In the current public development landscape, CLDN18.2 and HER2 are the two most clinically actionable candidates for a first esophageal adenocarcinoma dual-target CAR-NK concept.

This example therefore advances a CLDN18.2/HER2 tandem-target product and treats EGFR as a pre-specified exploratory biomarker, not as the primary CAR target in version 1.

Investigational product. EBNK-1822H2 is a hypothetical off-the-shelf NK-cell product derived from cord blood and engineered with a tandem CAR recognizing CLDN18.2 and HER2, together with a persistence-support module (for example, membrane-bound IL-15) and an inducible caspase-9 safety switch. This is an example investigational product description for protocol-drafting purposes only.

Study structure. Phase 1 uses an open-label dose-escalation design after fludarabine/cyclophosphamide lymphodepletion to identify the maximum tolerated dose (MTD) and recommended phase 2 dose/schedule (RP2D/RP2S). Phase 2 expands at the selected regimen in three biomarker-defined strata: (1) CLDN18.2-positive/HER2-negative, (2) HER2-positive/CLDN18.2-low or negative, and (3) dual-positive disease. Participants receive one intravenous infusion on Day 0; a protocol-defined repeat infusion on Day 21 may be allowed in the expansion portion if there is no DLT or uncontrolled grade 3+ immune toxicity.

Correlative program. Screening includes central or local assessment of CLDN18.2 and HER2, with EGFR captured prospectively at baseline and, when feasible, at progression. Correlative studies include ctDNA response, cytokine profiling, CAR-NK persistence, tumor microenvironment markers, and antigen-pattern analyses to inform future multi-target program evolution.

Inclusion Criteria:

• Histologically confirmed esophageal adenocarcinoma or Siewert I/II gastroesophageal junction adenocarcinoma judged biologically consistent with esophageal adenocarcinoma, unresectable/recurrent/metastatic, and not amenable to curative therapy.

• Disease progressed after at least 1 prior systemic regimen for advanced disease, or the participant is intolerant of / ineligible for standard therapy. Biomarker-directed therapy must have been received or deemed inappropriate/unavailable where standard in the local setting.

• At least 1 measurable lesion by RECIST v1.1.

• Evidence of at least one selected target: CLDN18.2-positive by validated IHC (example threshold: membranous staining in

  ≥75% of tumor cells with moderate/strong intensity or protocol-specified central threshold), and/or HER2-positive by IHC 3+ or IHC 2+/ISH-amplified disease.

• ECOG performance status 0-1.

• Adequate marrow, renal, hepatic, pulmonary, and cardiac function per protocol laboratory thresholds.

• Life expectancy ≥12 weeks.

• Resolution of clinically significant prior-therapy toxicities to grade ≤1 (except alopecia or stable endocrinopathies).

• Willingness to provide archival tumor tissue and to undergo fresh biopsy when safely feasible.

• Negative pregnancy test for participants of childbearing potential and agreement to protocol-defined contraception.

Exclusion Criteria:

• Esophageal squamous cell carcinoma or non-adenocarcinoma histology.
• Known active CNS metastases or leptomeningeal disease; previously treated stable CNS disease may be allowed only if asymptomatic and off escalating steroids per protocol.
• Prior gene-modified cellular therapy within 12 weeks, or another investigational therapy likely to confound interpretation of safety or efficacy.
• Active uncontrolled infection, including uncontrolled hepatitis B, hepatitis C, HIV viremia, sepsis, or clinically significant opportunistic infection.
• Ongoing systemic immunosuppressive therapy above physiologic steroid replacement.
• Active autoimmune disease requiring systemic treatment within 2 years.
• Clinically significant interstitial lung disease/pneumonitis, uncontrolled cardiovascular disease, or left ventricular ejection fraction <50%.
• Active gastrointestinal perforation, uncontrolled bleeding, or clinically significant mucosal ulceration that would increase study-treatment risk.
• Prior solid organ transplant or active graft-versus-host disease.
• Pregnant or breastfeeding.
• Another active malignancy requiring systemic therapy, except protocol-permitted low-risk cancers.

Masking is not used because this is an early-phase adoptive cell therapy study requiring real-time safety monitoring, dose-escalation decisions, and close coordination of cell-product logistics.