Pivekimab sunirine (IMGN632), a novel CD123-targeting antibody-drug conjugate, in relapsed or refractory acute myeloid leukaemia: a phase 1/2 study

Abstract

Background: Pivekimab sunirine (IMGN632) is a first-in-class antibody-drug conjugate comprising a high-affinity CD123 antibody, cleavable linker, and novel indolinobenzodiazepine pseudodimer payload. CD123 is overexpressed in several haematological malignancies, including acute myeloid leukaemia. We present clinical data on pivekimab sunirine in relapsed or refractory acute myeloid leukaemia.

Methods: This first-in-human, phase 1/2 dose-escalation and dose-expansion study enrolled participants aged 18 years or older at nine hospitals in France, Italy, Spain, and the USA with CD123+ haematological malignancies (Eastern Cooperative Oncology Group performance status of 0-1); participants reported here were in a cohort of participants with acute myeloid leukaemia who were refractory to or had relapsed on one or more previous treatments for acute myeloid leukaemia. The 3 + 3 dose-escalation phase evaluated two dosing schedules: schedule A (once every 3 weeks, on day 1 of a 3-week cycle) and fractionated schedule B (days 1, 4, and 8 of a 3-week cycle). The dose-expansion phase evaluated two cohorts: one cohort given 0·045 mg/kg of bodyweight (schedule A) and one cohort given 0·090 mg/kg of bodyweight (schedule A). The primary endpoints were the maximum tolerated dose and the recommended phase 2 dose. Antileukaemia activity (overall response and a composite complete remission assessment) was a secondary endpoint. The study is ongoing and registered with ClinicalTrials.gov, NCT03386513.

Findings: Between Dec 29, 2017, and May 27, 2020, 91 participants were enrolled (schedule A, n=68; schedule B, n=23). 30 (44%) of schedule A participants were female and 38 (56%) were male; 60 (88%) were White, six (9%) were Black or African American, and two (3%) were other races. Pivekimab sunirine at doses of 0·015 mg/kg to 0·450 mg/kg in schedule A was administered in six escalating doses with no maximum tolerated dose defined; three dose-limiting toxicities were observed (reversible veno-occlusive disease; 0·180 mg/kg, n=1 and 0·450 mg/kg, n=1; and neutropenia; 0·300 mg/kg, n=1). Schedule B was not pursued further on the basis of comparative safety and antileukaemia findings with schedule A. The recommended phase 2 dose was selected as 0·045 mg/kg once every 3 weeks. At the recommended phase 2 dose (n=29), the most common grade 3 or worse treatment-related adverse events were febrile neutropenia (three [10%]), infusion-related reactions (two [7%]), and anaemia (two [7%]). Treatment-related serious adverse events occurring in 5% or more of participants treated at the recommended phase 2 dose were febrile neutropenia (two [7%]) and infusion-related reactions (two [7%]). Among 68 participants who received schedule A, one death (1%) was considered to be treatment-related (cause unknown; 0·300 mg/kg cohort). At the recommended phase 2 dose, the overall response rate was 21% (95% CI 8-40; six of 29) and the composite complete remission rate was 17% (95% CI 6-36; five of 29).

Interpretation: Pivekimab sunirine showed single-agent activity across multiple doses, with a recommended phase 2 dose of 0·045 mg/kg once every 3 weeks. These findings led to a phase 1b/2 study of pivekimab sunirine plus azacitidine and venetoclax in patients with CD123-positive acute myeloid leukaemia.

Funding: ImmunoGen.

Figure 1:. Trial profile

No participants who were treated in the dose-escalation cohort were included in the dose-expansion cohort.*The starting dose for schedule A was 0·015 mg/kg; dose escalation followed a modified Fibonacci schema. Progressive disease was defined in the protocol (appendix). Clinical progression included participants who did not meet the criteria for progressive disease but had smaller increases in blasts or had circulating blasts and were taken off treatment. ‡The maximum starting dose for schedule B was no higher than the last dose cleared on schedule A, administered in equal amounts across days 1, 4, and 8 of a 3-week cycle (eg, a third of the total dose administered on schedule A on each of days 1, 4, and 8).

Figure 2:. Greatest decrease in bone marrow blasts in all schedule A participants

Responders included those in complete remission or complete remission with partial haematological recovery or incomplete recovery; or in a morphological leukaemia-free state; or in partial remission. Eight participants are not represented in the plot because two participants had clinical progression before the bone marrow assessment, and six participants discontinued because of adverse events before the bone marrow assessment. 11 participants had a more than 100% increase in bone marrow blasts. One participant had a greatest decrease in bone marrow blasts of −99·883% based on flow minimal residual disease. However, the morphological blast assessment was not reported in the same marrow assessment, which is why an International Working Group response could not be confirmed for this participant as per the protocol. After discussion, a best overall response of stable disease was eventually entered by the site in the electronic data capture system for this participant.