LILRB4 Synthetic T-Cell Receptor and Antigen Receptor-T (STAR-T) for Refractory/Relapsed Acute Myeloid Leukemia: First-in-Human Phase I Clinical Trial

Introduction CAR-T cell development for refractory or relapsed (R/R) acute myeloid leukemia (AML) has been limited by the lack of specific antigens with high specificity for AML cells and not present on normal hematopoietic stem cells. LILRB4 (also known as ILT3 or CD85k) is an immunosuppressive receptor and highly expressed on monocytic AML blasts (FAB M4 and M5 AML subtypes), which makes it an ideal therapeutic target for monocytic AML. Methods LILRB4-specific nanobodies were obtained from alpacas immunized with the recombinant LILRB4 extracellular domain. These nanobodies were then fused to IgG1 Fc, expressed and purified from CHO cells. The construct used for the manufacture of LILRB4-specific nanobody-based STAR-T (Synthetic T-Cell Receptor and Antigen Receptor-T) contains a double-chain TCR-based receptor with nanobodies fused to the N termini of modified TCR-Cα or TCR-Cβ and OX40 co-stimulatory domain (named Biparatopic LILRB4 STAR-T). In vitro assays and xenograft models were used to assess the potential of LILRB4 STAR-T cells for elimination of leukemic disease. In the first-in-human, single-arm, open-label phase I investigator-initiated trial (NCT05548088), we investigated the safety and efficacy of LILRB4 STAR-T therapy for treating R/R AML patients. By the Bayesian Optimal Interval (BOIN) design, patients who met the inclusion/exclusion criteria were enrolled to evaluate the safety, efficacy, and dose-limited toxicities (DLTs) by four dose groups (DL: 1E6, 3E6, 6E6, 1E7 STAR-T/kg). Peripheral blood mononuclear cells were obtained from either the patients themselves or the transplant donor post allo-HSCT. T-cells were then purified using CD3/CD28+ magnetic beads. All patients were treated with FC lymphodepletion before STAR-T cell infusion. Results In vitro and in vivo cytotoxicity assays demonstrated that LILRB4 STAR-T cells were highly cytotoxic against LILRB4+ AML cell lines, and also could inhibit the immunosuppressive myeloid cells (such as macrophage, and monocytes). As of July 31th, 2024, a total of 9 patients with LILRB4 -positive R/R AML were enrolled and evaluable for safety and efficacy. Median follow-up was 5 months (range 0.5-13.5months). The median age of enrolled patients was 36 years (range 25-60), and 77.8% (7/9) of patients relapsed post allo-HSCT. Median bone marrow (BM) blast percentage was 60% (12-81%) at enrollment to lymphodepletion. LILRB4 expression level in blasts was 77% (42-97%). The median transduction efficiency of the products was 54.75% (32.4%-75.5%). In the BOIN dose escalation, patients received STAR-T therapy at doses of 1E6/kg (n=1), 3E6/kg (n=3), 6E6/kg (n=4), and 1E7/kg (n=1). Six patients completed the 28-day dose-limiting toxicity (DLT) assessments. Two patients died due to infection before the d28 DLT assessments, and one patient's update will be provided at the ASH meeting. The safety analysis revealed that 83.3% (5 out of 6) of patients developed grade 1-2 cytokine release syndrome (CRS), with no grade 3 CRS or immune effector cell-associated neurotoxicity syndrome (ICANS) observed. The most common adverse event was pancytopenia, and no target-related adverse events were reported. Following infusion, the median peak time for circulating STAR-T cells occurred around Day 7 (range Day 5 to Day 10), as measured by q-PCR. The presence of LILRB4-positive cells in peripheral blood (PB) was negatively correlated with the level of STAR-T cells in PB. In responsive patients, LILRB4-positive AML cells in BM were significantly reduced and nearly disappeared. The efficacy data indicated a best overall response rate (ORR) of 50.0% (3 out of 6), comprising complete remission (n=1), morphologic leukemia-free state (n=1), and partial remission (n=1). Two of the responders underwent a second or third allo-HSCT and survived without minimal residual disease (MRD). The estimated 1-year OS was 62.2 % (95CI%: 35.5-100). Conclusion Our first-in-human study demonstrated that LILRB4 STAR-T therapy is a promising approach in LILRB4 -positive r/r AML patients, even in those who have undergone extensive prior treatments and experienced relapse post allo-HSCT. Further data from the phase II trial and longer follow-up time are essential.