New Macrophage Immune Checkpoint Strategy Shows Promise in AML and MDS

David Sallman, MD

Hu5F9-G4 (5F9), a first-in-class antibody targeting CD47, used as monotherapy or in combination with standard azacitidine was well tolerated and provided deep and durable responses in patients with acute myeloid leukemia (AML) or myelodysplastic syndrome (MDS), according to findings from a phase IB study presented at the 2019 European Hematology Association (EHA) Congress.

5F9 is a novel form of immunotherapy that targets CD47 to restore innate immunity, the first line of defense against cancer cells. CD47 provides the notorious “don’t eat me” signal displayed by many types of cancer cells that enables macrophage immune evasion by preventing phagocytosis. CD47 is the dominant macrophage checkpoint that is overexpressed on most cancer cells and increased CD47 expression has been associated with poorer prognosis.

David Sallman, MD, Moffitt Cancer Center, explained the mechanism behind the combination treatment. “Azacitidine induces pro-phagocytic “eat me” signals on cancer cells, which acts synergistically with the CD47 blockade provided by the antibody overriding the “don’t eat me” signal and leading to enhanced phagocytosis.”

Therefore, the 5F9005 study (NCT03248479) was designed to compare the activity of 5F9 administered alone or in combination with azacitidine in patients with AML and patients with MDS. In addition to patients with relapsed/refractory AML, the study enrolled patients with previously untreated AML who were ineligible for induction chemotherapy and untreated MDS patients at intermediate to very high risk by IPSS-R criteria. Ten patients with relapsed/refractory AML received sole 5F9 at 30 mg/kg twice weekly and 36 treatment-naive patients were treated with the standard regimen of azacitidine (75 mg/m² on days 1 to 7) plus a priming dose of 5F9 of 1 mg/kg followed by ramp-up to 30 mg/kg given weekly. The priming dose was used to mitigate on-target anemia.

“On target anemia is a pharmacodynamic effect that is mitigated with a 5F9 priming and maintenance dosing regimen,” he explained. “Many patients have had hemoglobin improvement and decrease in transfusion frequency while on 5F9 and azacitidine.”

In the cohort of patients with relapsed /refractory disease, 6 patients with AML and 4 with MDS received 5F9 monotherapy. These patients had received a median of 2 (range, 1-6) prior therapies.

In the cohort of patients with untreated AML or MDS, 5F9 was administered with azacitidine as first line to 22 patients with AML and 17 patients with MDS. In the respective combination therapy arms, the median patient age was 74 and 71 years. Fifty-five percent of patients with AML also had underlying myelodysplasia (MRC). Among MDS patients, 41% of patients were at high risk and 12% of patients were at very high risk.

The primary objectives of the study were to evaluate the safety and efficacy of 5F9 treatment in relapsed/refractory AML and 5F9 in combination with azacitidine in patients with untreated AML or MDS.

The safety analysis showed 5F9 was well tolerated both as monotherapy (n = 10) and in combination (n = 36). The maximum tolerated dose was not reached in either treatment arm and the safety profile of 5F9 combined with azacitidine was similar to that of single agent azacitidine. In the combination arm one dose-limiting toxicity (DLT), grade 4 hemagglutination, was observed that resolved within 24 hours. One patient discontinued combination treatment due to an adverse event (AE). No cytopenias, infections, or autoimmune AEs were observed and no deaths occurred within the initial 60 days of treatment.

Antileukemic activity was observed with 5F9 monotherapy and in combination in patients with AML and MDS. Among the 10 patients in the relapsed/refractory AML cohort receiving 5F9 monotherapy the objective response rate was 10%, which represented a marrow complete response (CR); 70% of patients demonstrated stable disease (SD) and 20% experienced disease progression (PD).

In the cohort of 14 patients with AML and 11 patients with MDS receiving 5F9 plus azacitidine as first line therapy, the ORR was 64% and 100% respectively.

In the AML subgroup, the responses included 5 (36%) patients with CR, 2 (14%) patients with morphologic complete remission with incomplete blood count recovery, and 2 (14%) patients with marrow CR. SD was seen in 36% of patients and no patients had PD.

Of the 11 responding patients with MDS, 6 (55%) patients had CR, 4 (36%) had marrow CR, and one (9%) patient showed hematologic improvement. Again, no disease progression occurred.

The time to response was more rapid at 1.9 months with the combination that that observed with azacitidine alone.

“The responses with combined 5F9 and azacitidine were deep and durable,” Sallman remarked.

Red blood cell (RBC) transfusion independence was achieved by 64% of patients with AML and 29% of responding patients with AML showed a complete cytogenetic response. MRD negativity was seen in 33% of responders. The median duration of response (DoR) was not reached (range, 0.03+ to 8.3+) months.

In patients with MDS treated with the combination, 43% had a complete cytogenetic response, and 20% showed MRD negativity. Median DoR was not reached (range, 0.5+ to 4.3+).

The median follow-up in both groups was short, approximately 3.8 months, and is ongoing.

“The longest patient in response is in CR for 9+ months in therapy and ongoing,” Sallman noted.

“Combination treatment also eliminated leukemic mutations in some AML patients and eliminated leukemia stem cells in responding AML and MDS patients,” Sallman said.

He also noted that “Extension cohorts have been initiated in AML and MDS.”

Sallman D, Donnellan W, Asch A. The first-in-class anti-CD47 antibody HU5F9-G4 is active and well tolerated alone or in combination with azacitidine in AML and MDS patients: initial phase Ib results. Presented at 2019 EHA Congress; June 13-16, 2019; Amsterdam, The Netherlands. Abstract S878.

<<< 2019 European Hematology Association Congress