Phase 1/2 study evaluates the safety, feasibility, and preliminary antitumor activity of allogeneic dual-target CD70/CAIX CAR-NK cells after fludarabine/cyclophosphamide lymphodepletion in adults with advanced or metastatic clear cell RCC that has progressed after standard therapy. The study is designed to determine a recommended dose and schedule, characterize hepatobiliary safety, and explore whether CD70-high, CAIX-high, or dual-high tumors derive the greatest benefit. Biomarker-defined activity signals will be used to guide whether later development should prioritize CD70, CAIX/CA9, or continued dual-targeting.

Advanced clear cell RCC remains difficult to treat after immune checkpoint blockade and VEGF-pathway therapy. Both CD70 and CAIX are attractive RCC targets: CAIX is broadly expressed in clear cell RCC, while CD70 is frequently overexpressed and has already generated encouraging clinical cell-therapy signals in RCC [3-11]. The investigational product in this example is an allogeneic, cord blood-derived NK cell product engineered to express a dual CAR recognizing CD70 and CAIX/CA9, membrane-bound IL-15 to enhance persistence, and an inducible caspase-9 safety switch to improve controllability. The allogeneic CAR-NK strategy is intended to provide an off-the-shelf platform with lower theoretical risk of graft-versus-host disease than allogeneic T-cell products and with innate NK killing as a complementary antitumor mechanism [1, 2, 11]. The study uses a standard dose-escalation phase followed by a biomarker-guided expansion phase. All participants receive lymphodepletion with fludarabine and cyclophosphamide and then one infusion of dual-target CAR-NK cells on Day 0; an optional second infusion on Day 15 is allowed for participants without doselimiting toxicity, uncontrolled cytokine-mediated toxicity, or rapid progression.

During expansion, participants remain in a single treatment arm but are prospectively analyzed in CD70-high, CAIX-high, and dual-high subgroups. Because earlier CAIX CAR-T studies in metastatic RCC reported on-target hepatobiliary toxicity related to low-level CAIX expression in biliary epithelium, this example protocol incorporates enhanced liver screening, exclusion of significant biliary disease, staggered first-patient dosing at each dose level, frequent liver function monitoring, and an explicit early stopping rule for recurrent protocoldefined hepatobiliary toxicity [5, 6].

The translational objective is not only to identify a recommended phase 2 dose but also to learn which antigen context appears most actionable in patients. If activity clusters in CD70-high tumors with cleaner safety, later development may pivot toward a CD70-led construct. If CAIX-high tumors uniquely benefit without meaningful hepatobiliary toxicity, CAIX/CA9 may remain central. If dual-high tumors clearly outperform single-high tumors without added risk, the dual-target program should continue.

Inclusion Criteria:

• Written informed consent and willingness to comply with protocol procedures.
• Age >= 18 years at the time of consent.
• Histologically confirmed unresectable or metastatic clear cell RCC, or RCC with a clear-cell component, with radiographic progression after standard therapy.
• Prior exposure to at least one PD-1 / PD-L1-based regimen and at least one VEGF-pathway targeted regimen, or documented intolerance / unsuitability for available standard systemic options.
• At least one measurable lesion by RECIST 1.1.
• Available archival tumor tissue or willingness to undergo fresh biopsy for central biomarker testing; protocoldefined tumor positivity for CD70 and/or CAIX is required. Dual-positive cases are preferred for the biomarkerexpansion portion.
• ECOG performance status 0-1.
• Adequate marrow, liver, cardiac, pulmonary, and renal function as defined by the protocol (for example, ANC, platelets, bilirubin, AST/ALT, creatinine clearance, oxygen saturation, and left ventricular function within protocoldefined limits).
• Life expectancy of at least 12 weeks.
• Negative pregnancy test for participants of childbearing potential and agreement to highly effective contraception during protocol-defined risk windows.
• Previously treated brain metastases are allowed if clinically stable and off escalating corticosteroids for at least 14 days before lymphodepletion.

Exclusion Criteria:

• Active, untreated, or symptomatic central nervous system metastases, leptomeningeal disease, or uncontrolled seizure disorder.
• Prior gene-modified cellular therapy (including prior CAR-T, CAR-NK, CAR-NKT, or TCR-engineered therapy) within the protocol-defined washout window.
• Prior allogeneic stem cell transplant or solid organ transplant with ongoing clinically significant immunosuppression.
• Active autoimmune disease requiring systemic immunosuppressive treatment; physiologic replacement doses are permitted.
• Active uncontrolled infection, including uncontrolled hepatitis B, hepatitis C, HIV, tuberculosis, or sepsis.
• Clinically significant hepatobiliary disease that could increase risk from CAIX-directed therapy, such as active cholangitis, primary sclerosing cholangitis, biliary obstruction, Child-Pugh B/C cirrhosis, or prior hepatic venoocclusive disease.
• Clinically significant cardiovascular disease (for example, unstable angina, recent myocardial infarction, uncontrolled arrhythmia, or clinically meaningful heart failure).
• Systemic corticosteroid use greater than 10 mg prednisone equivalent daily within 7 days before lymphodepletion, unless required as physiologic replacement.
• Pregnancy or breastfeeding.
• Another active invasive malignancy requiring systemic treatment, except for protocol-defined low-risk exceptions.
• Known hypersensitivity to fludarabine, cyclophosphamide, or a critical study-product excipient.

The trial is open-label because the main objectives are dose finding, real-time toxicity review, cell-product release management, and intensive translational monitoring.