Dose Escalation of HLA-A2-WT1 CD3 T-Cell Bispecific Antibody in a Phase I Study in Patients with Relapsed/Refractory Acute Myeloid Leukemia (AML)

Background: RO7283420 (RG6007) is a 2+1 TCR-like (TCR-L) T-cell bispecific (TCB) antibody targeting CD3 and the RMFPNAPYL peptide of Wilms Tumor 1 (WT1) protein presented by the major histocompatibility complex-I HLA-A*02 on acute myeloid leukemia (AML) blasts and other antigen presenting cells. The pre-clinical evaluation of RO7283420 in our in vivo humanized AML xenografts and ex vivo AML co-culture models showed strong T-cell mediated AML cell killing (Augsberger C, et al. Blood 2021). A Phase 1 dose-escalation (DE) study (NCT04580121) evaluated the safety, tolerability, pharmacokinetics, anti-drug antibodies (ADAs), and anti-leukemic activity of RO7283420 in patients with relapsed/refractory (R/R) AML.

Methods: In this open-label, multi-center study, DE was performed using a 3+3 design. Patients received RO7283420 every 3 weeks (Q3W, n=46), or every week (QW, n=4) as intravenous (IV) infusions. Preliminary anti-leukemic activity included response assessment according to European Leukemia Net (ELN) response categories (adapted from Döhner H, et al. Blood 2017).

Results: As of 13 April 2023, 50 HLA-A2+ R/R AML patients received at least one dose of RO7283420. Patients had a median age of 65.5 years (range 35-84), presented with ECOG of 0 (56%) or 1 (38%) or 2 (6%), and 58% were male. Median prior line of therapy was 2 (range 1–5). Overall, 58% of patients had relapsed and 42% had primary refractory disease. According to the ELN 2017 risk stratification, patients were of adverse (48%), intermediate (38%) or favorable (8%) risk category, for 6% of patients the risk category missing. The most common genetic abnormalities reported were RUNX1 (21%), ASXL1 (17%), TP53 (10.6%), FLT3-ITD (6.4%) and NPM1 (6.4%) of the 47 patients tested. Median bone marrow blast percentage at baseline was 35% (range 3–90%), while the median of circulating blast was 17% (range 0-88%). Study patients received RO7283420 IV at 13 different dose levels, ranging from a Minimum Anticipated Biological Effect Level (MABEL) of 0.15 mg (flat) to 18 mg Q3W (with a preceding weekly 1/3 mg ‘double step-up’ during Cycle 1) and one QW dose level with 9 mg preceded by a 1/3 mg step-up. Maximum tolerated dose was reached at 1/3/12 mg double step-up Q3W. Explored alternative QW schedule was not tolerable at 1/3/9 mg.

The most common (≥20%) adverse events (AEs) were cytokine release syndrome (CRS) occurring in 34 (68%), pneumonia 14 (28%), pyrexia 13 (26%), febrile neutropenia 13 (26%), hyperglycemia 12 (24%), hypokalemia 11 (22%), and nausea 10 (20%) of patients. Eight dose limiting toxicities (DLTs) were reported: five G3 CRS (at 1/3, 2/12, 1/6/12 mg dose levels), G3 stomatitis (at 1/3/12 mg), G3 myositis (at 1/3/18 mg), and G4 thrombocytopenia (at 1/3/18 mg). Eleven patients (22%) experienced Grade 5 AEs with pneumonia, sepsis, and hemophagocytic lymphohistiocytosis (HLH) reported in >1 patient; all Grade 5 events were considered unrelated to RO728342 except 1 (at 1/3/9 mg) of 2 HLH events.

IV PK has been overall dose-linear and characterized by a terminal half-life of 29 to 84 hours and a clearance of 58 to 92 mL/h. Preliminary ADA incidence within the study population was 19%. In the dose ranges tested, a trend for study drug exposure-dependent blast reduction was observed in blood, while a clear exposure-response relationship could not be established with BM blast reduction. Furthermore, our preliminary pharmacodynamic analysis identified expansion of naive and memory CD8 T cells in blood and activated CD8+ T cells in bone marrow, in line with the expected mode of action (MoA) of RO7283420. Preliminary efficacy signals were observed with 3 complete responses (CR), including 1 CR with incomplete blood count recovery.

Conclusions: RO7283420 is the first TCR-L TCB antibody evaluated in AML. We observed pharmacodynamic evidence of T-cell activation and expansion in the clinic, in line with the expected MoA of TCBs, however, at the explored doses, no clear exposure-response relationship and only a modest clinical activity were observed. The safety profile was shown to be consistent with the other TCBs and R/R AML population. Based on the totality of data, the study was discontinued.