Evorpacept (ALX148), a CD47-Blocking Myeloid Checkpoint Inhibitor, in Combination with Azacitidine: A Phase 1 / 2 Study in Patients with Myelodysplastic Syndrome (ASPEN-02)

Background: Evorpacept is a high affinity CD47-blocking fusion protein with an inactive human immunoglobulin Fc region designed to enhance the activity of other anti-neoplastic therapies, such as azacitidine (AZA), with minimal additional toxicity. Here, we present results from the phase 1 part of the ASPEN-02 study evaluating the safety and tolerability of evorpacept administered in combination with AZA in subjects with myelodysplastic syndrome (MDS).

Methods: ASPEN-02 is a phase 1/2 open-label, multicenter study designed to evaluate the safety and tolerability and establish the recommended phase 2 dose (RP2D) of intravenous evorpacept in combination with AZA in phase 1 and to evaluate the efficacy of the combination of evorpacept + AZA compared to AZA alone in subjects with previously untreated higher risk MDS in phase 2. Adult subjects with newly diagnosed (ND) higher risk (IPSS-R>3.5) or relapsed/refractory (R/R) MDS were enrolled into phase 1 cohorts receiving escalating doses of evorpacept (20 mg/kg Q2W, 30 mg/kg Q2W, and 60 mg/kg Q4W) combined with AZA (75 mg/m ² IV/SC x 7d) in a 28-day treatment cycle. The primary phase 1 endpoint is the frequency of first cycle dose-limiting toxicities (DLTs). The study is open for enrollment (NCT: NCT0441751).

Results: As of July 15, 2021, 13 subjects were treated in phase 1 at evorpacept doses of 20 mg/kg Q2W (N=3), 30 mg/kg Q2W (N=3), and 60 mg/kg Q4W (N=7). Of the 7 ND subjects, 4 had therapy-related MDS, and 5 had TP53 mutation with complex cytogenetics. Of the 6 R/R subjects, all had received at least 1 prior hypomethylating agent (HMA)-based regimen. Median age of the phase 1 population was 74 years (range 56-82), and baseline ECOG scores were 0 (n=4) or 1 (n=9).

No DLTs were observed in any cohort and a maximum tolerated dose (MTD) was not reached. All subjects experienced an adverse event. Treatment-related AEs (TRAEs) observed in >1 subject included constipation and infusion related reaction (n=3 each, 23%), and nausea and vomiting (n=2 each, 15%). Grade 3 or higher AEs of any causality occurring in >1 subject included febrile neutropenia (n=4; 31%), pneumonia (n=3, 23%), and anemia and thrombocytopenia (n=2 each, 15%). There was one Grade 3 or higher TRAE of transient neutropenia reported. There were no evorpacept-related serious adverse events (SAEs), no deaths on study, and no patients discontinued treatment due to an AE. Preliminary PK and PD data indicated dose-proportional pharmacokinetics consistent with prior studies, and evidence of full CD47 target occupancy (TO) in peripheral blood at both evorpacept peak and trough concentrations when administered every 2 or 4 weeks.

Among the 5 ND subjects evaluable for response (all with TP53 mutation), there was 1 subject with a best response of marrow CR with hematologic improvement (HI), 2 subjects with cytogenetic response including 1 with HI, and 1 subject each with SD and PD. Of the 4 ND subjects who were transfusion dependent at baseline, 2 achieved transfusion independence. Among the 5 R/R subjects evaluable for response, there were 2 subjects with a best response of marrow CR, 2 with SD, and 1 with PD. Additional results will be presented at the meeting.

Conclusions: The myeloid checkpoint inhibitor evorpacept was well tolerated in combination with AZA with an initial safety profile similar to that of AZA monotherapy. The evorpacept RP2D in combination with AZA is 60 mg/kg Q4W. Promising initial activity has been observed, with objective responses and hematologic improvement at multiple dose levels in subjects with both ND and R/R MDS, including those with adverse risk features such as TP53 mutation with complex cytogenetics, therapy-related MDS, and prior HMA-failure. This combination will be evaluated further in the randomized phase 2 part of this study.