New Hope for Antibody-based Immunotherapy for AML: Will the DART Hit the Bullseye?

Immunotherapy in the form of allogeneic hematopoietic stem cell transplantation (allo HSCT) has been a foundation of curative therapy in acute myeloid leukemia (AML) since the 1970s when Dr. E. Donnall Thomas and colleagues reported that patients with refractory acute leukemia could achieve long-term leukemia-free survival with high-dose chemotherapy followed by human leukocyte antigen (HLA)–identical sibling marrow transplantation.^1-4  This approach has been successful, and significant advances have decreased transplant-related mortality (TRM) during the past 40 years due to the use of better drugs for prophylaxis of graft-versus-host disease and infections, reduced intensity or nonmyeloablative conditioning (RIC) regimens, improved HLA matching, and the development of the comorbidity index; however, relapse after transplant remains the major cause of treatment failure in AML and substantial TRM (2-year TRM is 25-39% in adults aged >40 years with AML)⁵  remains a barrier to transplant utilization in AML. Accordingly, nontransplant immunotherapy approaches such as targeted antibody-based therapy and chimeric antigen receptor T-cell therapy, which are exceedingly successful in acute lymphoblastic leukemia, have been enthusiastically pursued as potential therapeutic strategies to harness the potency of cell-based immune therapy, but with greater target selectivity in AML. The development of such therapies, however, has been limited to the approval of only a single antibody-targeted therapy for AML, the anti-CD33 antibody-drug conjugate gemtuzumab ozogamicin.⁶  The major hurdle to the development of antibody-based therapy is related to the few suitable cell surface antigen targets that can differentiate AML blasts from normal hematopoietic progenitor cells.

CD123, the interleukin 3 (IL-3) receptor α chain (IL3RA), is a promising tumor-associated antigen target for AML and one that appears amenable to immunotherapy due to its high expression on the surface of AML blasts and leukemia stem cell populations,⁷  as well as its limited expression on normal hematopoietic stem and progenitor cells and other normal myeloid tissues (i.e., plasmacytoid dendritic cells, monocytes, and basophils).⁸  Early attempts of therapeutics targeting CD123 included a diphtheria toxin IL-3 fusion protein⁹  and a monoclonal antibody to CD123 (CSL-360)¹⁰  and demonstrated antileukemia activity in vitro and/or in vivo, but when tested in phase I clinical trials, both agents lacked clinical activity in AML. In contrast, flotetuzumab, a drug that has shown promising preclinical activity, is an investigational dual affinity retargeting (DART) antibody that targets both CD123-positive AML cells and CD3-expressing T lymphocytes to form an artificial immunologic synapse, and then induces T-cell activation, T-cell expansion, and T-cell–mediated CD123 targeted killing.^11-13 

In the current article, Dr. Geoffrey Uy and colleagues report the results of a first in-human phase I/II single arm, multicenter study of flotetuzumab in 88 patients with relapsed or refractory (R/R) AML. The primary objective of the study was to determine the maximum tolerated dose (MTD) and schedule of flotetuzumab and to characterize the toxicity profile. Secondary objectives included pharmacokinetic and pharmacodynamic profile and the antitumor activity of flotetuzumab. The trial enrolled 42 patients to the phase I dose-finding cohort, and 46 patients were enrolled onto the phase II dose-expansion cohort. The recommended phase II dose (RP2D) of flotetuzumab is 500 ng/kg/day administered as a continuous infusion in 28-day cycles, following a "lead-in dose" phase during week 1 of treatment. Ninety-six percent of patients treated at the RP2D experienced infusion-related reactions/cytokine release syndrome (IRR/CRS), though the majority (81%) had mild to moderate signs/symptoms (grade 2). Seven patients (8%) experienced grade 3 IRR/CRS, and no grade 4/5 events occurred. Most IRR/CRS events (32%) occurred during the first week of treatment with the "lead-in dose" phase, and the IRR/CRS events decreased throughout the 28-day cycle. The IRR/CRS events were managed with drug interruptions and/or tocilizumab. Neurologic events were infrequent overall, with the most common event being headache (grade 1 or 2 in 10%). Two patients experienced grade 3 neurologic events: one headache and one episode of delirium. Both patients recovered from these events. Importantly, flotetuzumab at the RP2D did not appear to result in prolonged myelosuppression. The overall complete response rate (ORR) for the RP2D was 24 percent (12 of 50), and the complete response with or without complete hematologic recovery (CR/CRh) was 18 percent (9 of 50). The median time-to-first response was 0.84 months (range, 0.8-2.1 months), and median overall survival (OS) was 3.2 months (95% CI, 2.10-6.47).

Exploratory subgroup analysis revealed that patients with primary induction failure (PIF) or early relapse occurring within six months after initial treatment (ER) had higher response rates (43%; 12 of 28) than patients with late relapse (LR; 14%; 1 of 7) or prior treatment with a hypomethylating agent (30%; 3 of 10). The six- and 12-month survival rates for patients with PIF/ER treated at the R2PD were 42 percent and 20 percent, respectively.

Previously, the authors described the immune architecture of AML marrow specimens from pediatric and adult patients and identified distinct immune subtypes — immune-infiltrated and immune-depleted — and reported significant differences in the immune gene expression profiles. Importantly, these subtypes were predictive of the response and/or resistance to cytotoxic chemotherapy and immune targeted therapy.¹⁴  In the current study, the authors extended their prior results and examined the role of immune gene-signatures and response to flotetuzumab in a subgroup of 38 patients. Specifically, they confirmed that patients with the immune-infiltrated subtype were less likely to have responded to cytotoxic chemotherapy (i.e., patients met criteria for PIF/ER), but more likely to respond to immunotherapy. Consistent with this observation, patients with PIF/ER showed higher tumor immune infiltration scores when compared to patients with LR, and response to flotetuzumab was correlated with high immune infiltration scores. Additionally, the authors identified a 10 gene-expression signature that was associated with CR to flotetuzumab, and the genes in the signature reflected an immune infiltrated marrow microenvironment.