TANKeR-70: TGF-β-receptor type 2 knockout allogeneic NK cells for renal cell carcinoma expressing CD70.

Abstract

TPS571 Background: Clear cell renal cell carcinoma (ccRCC) frequently expresses CD70. Chimeric antigen receptor (CAR) natural killer (NK) cells may have a more favorable toxicity profile than CAR T cells and warrant further exploration as a novel cancer treatment. Transforming growth factor beta receptor 2 (TGFβR2) is expressed on immune effector cells and its ligand, transforming growth factor-beta (TGFβ) has been associated with suppression of antitumor immune cell activity in longitudinal biopsies of patients with ccRCC treated with triple immunotherapy using nivolumab + ipilimumab + anti-VEGF TKI (Nat Commun. 2025; PMID: 39794332). The MD Anderson Institute for Cell Therapy Discovery and Innovation developed TGFBR2 KO CAR.70/IL-15 NK cells, an off-the-shelf umbilical cord-derived IL-15-transduced CAR NK cell product targeting CD70 with knockout of TGFβR2 to decrease the immunosuppressive effects of TGFβ in the tumor microenvironment. We hypothesize that TGFBR2 KO CAR.70/IL-15 NK cells will be a safe, tolerable, and feasible treatment for patients with treatment refractory ccRCC. Methods: This is a Phase I study testing the safety and efficacy of TGFBR2 KO CAR.70/IL-15 NK cells in metastatic ccRCC refractory to immune checkpoint and tyrosine kinase inhibitors. Expression of CD70 ≥10% by immunohistochemistry is required for enrollment. The primary objective is to determine the safety, tolerability, maximum tolerated dose and optimal cell dose. Secondary endpoints are objective response rate per RECIST v1.1, progression-free survival, and overall survival. Exploratory endpoints include CAR NK cell persistence; longitudinal changes in serum IL6, IFNγ, and TNFα; and changes in tumor tissue molecular features pre- and post-treatment. Patients will receive lymphodepleting chemotherapy with dexamethasone (to facilitate engraftment) followed by fludarabine 30 mg/m 2 /day and cyclophosphamide 500 mg/m 2 /day (days -5 to -3), with a single CAR NK infusion given on day 0. The dose-escalation phase will use the Bayesian optimal interval (BOIN) design assess up to 4 dose levels (ranging from 1.3 x 10^7 up to 4.0 x 10^8 cells) with N = 18 maximum total patients. Toxicity, efficacy, and NK cell engraftment kinetics will be used to determine the optimal cell dose which will subsequently be tested in up to 12 additional patients in the dose-expansion phase to further collect toxicity and preliminary efficacy data. Accrual to this study at MD Anderson opened in October 2025. Clinical trial information: NCT07072234 .