First-in-Human Trial Shows AMG 330 Is Tolerable in Relapsed/Refractory AML

AMG 330, an anti-CD33 bispecific T-cell engager (BiTE®) antibody construct, was generally tolerable and demonstrated anti-leukemic activity in patients with relapsed/refractory acute myeloid leukemia (AML), according to results from a first-in-human trial presented at the 2018 ASH Annual Meeting.

The researchers also found that step-up dosing, pretreatment with corticosteroids, and the administration of tocilizumab and intravenous (IV) fluids also was an effective strategy for mitigating the risk of cytokine release syndrome (CRS) – a major expected toxicity of this type of treatment.

“Bispecific T-cell engager antibodies, as the term implies, have two targets – one on the surface of the leukemic cells and one on the surface of the T cells,” lead study investigator Farhad Ravandi-Kashani, MBBS, from the University of Texas MD Anderson Cancer Center, told ASH Clinical News. These agents have a different mechanism of action [from those of] prior antibodies, he continued. “They work by bringing the T cells to close proximity of the leukemic cells, allowing T-cell–mediated killing of the leukemic cells, so BiTEs rely more on the patient’s immune system to kill the leukemic cells.”

In this phase I, dose-escalation study, the researchers sought to establish the optimal dosing schedule of AMG 330 (which is designed to bind CD33 on leukemia cells and CD3 proteins on T cells) in adults with AML that was relapsed and/or refractory to at least one treatment course.

Participants began treatment with “an extremely low dose of the drug,” Dr. Ravandi-Kashani said, “so, in the initial cohorts, we did not expect to see much activity.” The study included 12 dosing cohorts: who received AMG 330 at 0.5 μg to 4 μg/day as a continuous IV infusion; subsequent cohorts comprised three to six patients who received AMG 330 10 μg to 240 μg/day.

Dosing followed a stepwise approach: If patients completed the first treatment cycle without experiencing dose-limiting toxicity, they received up to five additional treatment cycles, until clinical benefit or futility.

At the time of presentation, data were evaluable for 40 patients, who had received a median of four prior treatments (range = 1-15 treatments).

A primary outcome of interest in this study was response to therapy, which was defined by the revised International Working Group criteria, as well as complete response (CR) with incomplete hematologic recovery.

Participants received a median of one cycle (range = 1-6 cycles) of AMG 330. Most patients (n=35; 87.5%) discontinued therapy, most commonly due to disease progression (n=27). Six patients discontinued because of adverse events (AEs) and two discontinued per request. Only three patients (7.5%) completed five cycles of treatment.

Twenty-nine patients (73%) experienced serious AEs, and were considered related to AMG 330 treatment. The most frequently reported serious AEs (observed in >1 patient) included:

  • CRS (n=11)
  • febrile neutropenia (n=7)
  • pneumonia (n=4)
  • leukopenia (n=4)
  • pyrexia (n=3)
  • thrombocytopenia (n=3)
  • subdural hematoma (n=2)

For patients who developed CRS, the researchers noted, the administration of corticosteroids, vasopressors, and IV fluids, as well as the interruption of AMG 330, resolved CRS signs and symptoms in an average of one day.

One patient died on study due to disease progression, and another patient died due to intracranial hemorrhage; however, neither death was deemed related to the study drug.

The initial target dose (480 μg/day) was associated with dose-limiting toxicities, including grade 2 CRS and grade 4 ventricular fibrillation, leading investigators to reduce the target dose to 240 μg/day. “In the latter cohorts, we have seen CRs,” Dr. Ravandi-Kashani reported, including two patients with CRs at a target dose of 240 μg/day, with a lead-in dose of 10 μg/day to 60 μg/day.

“One of the concerns with any drug that targets a marker that also is expressed on normal myeloid cells is that it can cause … cytopenias, but I think the occurrence of two CRs in latter cohorts will alleviate that concern, at least for now,” he noted.

These results indicate that stepwise dosing of AMG 330 “is a viable strategy” in relapsed/refractory AML, Dr. Ravandi-Kashani concluded. “We hope that with an increased number of patients treated and further adjustment of doses we can get much higher responses.”

Limitations of the analysis include its lack of a randomized design, lack of a placebo control or treatment comparator group, and the small number of patients in the current cohort.

The authors report financial relationships with Amgen, the manufacturer of AMG 330 and the sponsor of this trial.


Ravandi D, Stein AS, Kantarjian HM, et al. A phase 1 first-in-human study of AMG 330, an anti-CD33 bispecific T-cell engager (BiTE®) antibody construct, in relapsed/refractory acute myeloid leukemia (R/R AML). Abstract #25. Presented at the 2018 ASH Annual Meeting, December 1, 2018; San Diego, CA.