Immune ‘Don’t Eat Me’ Blocker Plus Azacitidine Shows 72.7% Response Rate in Phase II CMML Trial

A single-arm phase II trial has reported that timdarpacept (IMM01), a recombinant signal regulatory protein α–IgG1 fusion protein, combined with the hypomethylating agent azacitidine produced an overall response rate of 72.7% (16 of 22 efficacy-evaluable patients; 95% CI 49.8–89.3) when given as first-line therapy to adults with newly diagnosed higher-risk chronic myelomonocytic leukemia (CMML). The study, conducted at 13 hospitals in China and published in Experimental Hematology & Oncology, enrolled 24 treatment-naïve patients between August 4, 2022, and May 12, 2023, with a data cut-off of May 13, 2025.

CMML is a rare clonal hematopoietic stem cell disorder that shares features of myelodysplastic syndromes and myeloproliferative neoplasms, marked by persistent monocytosis and bone marrow dysplasia. Prognosis is generally poor, with a median overall survival of 12–29 months and a 15–20% risk of transformation to acute myeloid leukemia within 3–5 years. Allogeneic hematopoietic stem cell transplantation remains the only curative option, but advanced age, comorbidities, and procedure-related risks limit its use to roughly 10% of patients, leaving hypomethylating agents such as azacitidine as the only US Food and Drug Administration–approved drug therapies for the disease.

Existing hypomethylating agent therapy delivers modest results. Retrospective studies and clinical trials report overall response rates of 40–50% and complete response rates below 20% in HMA-treated CMML patients, with median survival of 12–18 months. A meta-analysis of azacitidine monotherapy cited in the study yielded an overall response rate of 45% (95% CI 40–51%) and a complete response rate of 16% (95% CI 10–24%). Although these agents can epigenetically restore hematopoiesis in a subset of responders, mutational allele burdens generally persist and disease progression is ultimately inevitable, creating a clear rationale for combination approaches.

The rationale for pairing IMM01 with azacitidine rests on the CD47–SIRPα axis, an immune checkpoint that lets tumor cells evade macrophage-mediated clearance. CD47 acts as a “don’t eat me” signal, and preclinical work has shown that azacitidine can increase pro-phagocytic “eat me” signals such as cell-surface calreticulin on myeloid leukemia cells, enhancing the effect of CD47–SIRPα blockade. IMM01 is a next-generation SIRPα-Fc fusion protein that simultaneously blocks the CD47 inhibitory signal and engages activating Fcγ receptors to drive antibody-dependent cellular phagocytosis. Notably, it does not bind human red blood cells or induce hemagglutination, a design feature intended to avoid the antigen-sink effect and anemia-related toxicity that have hampered earlier CD47 antibodies such as magrolimab.

The enrolled population carried high-risk features. The median age was 62.0 years (IQR 58.0–68.5; range 38–75), with 15 of 24 patients (62.5%) male. Half were classified as CMML-1 and half as CMML-2; 58.3% had myeloproliferative disease, and under the CPSS-mol molecular prognostic scheme, 66.7% were high risk and 33.3% intermediate-2 risk. Baseline blood counts were poor, with median hemoglobin of 69.5 g/L and median platelet count of 73.5 × 10⁹/L. Next-generation sequencing identified ASXL1 mutations in 59.1% of patients, RUNX1 in 45.5%, SRSF2 in 40.9%, and TET2 in 36.4%, among other recurrent alterations.

Among the 22 efficacy-evaluable patients—two were excluded after early withdrawal due to compliance issues before the first response assessment—the 72.7% overall response rate included 6 complete responses (27.3%), 3 marrow complete responses with hematologic improvement (13.6%), 6 marrow complete responses alone (27.3%), and 1 hematologic improvement alone (4.5%). Three patients (13.6%) had stable disease as their best response and three (13.6%) progressed. The median time to response was 1.8 months (range 0.9–3.6), and the median duration of response was 16.9 months (95% CI 5.1–not estimable), with durable response rates of 67.2% at 12 months and 26.9% at 18 months. Hematologic improvement was observed in 8 patients (38.1%), and of the 6 patients who were red blood cell transfusion-dependent at baseline, 3 achieved transfusion independence for at least 8 consecutive weeks.

Survival outcomes were encouraging but carry the uncertainty inherent to single-arm data with limited follow-up. With a median follow-up of 26.9 months, median progression-free survival was 17.8 months (95% CI 5.3–NE) and median overall survival was 22.2 months (95% CI 9.0–NE). By the data cut-off, 13 of the 22 efficacy-evaluable patients (59.1%) had died, most commonly from disease progression with transformation to acute myeloid leukemia. Exploratory, post hoc subgroup analyses found no significant differences in median progression-free or overall survival between CMML-1 and CMML-2 subtypes, and the authors explicitly caution that these comparisons were underpowered, unadjusted for multiplicity, and should not be interpreted as definitive.

Safety findings reflected the myelosuppressive nature of the regimen. Among all 24 patients, 95.8% experienced at least one treatment-related adverse event, and grade ≥ 3 treatment-related adverse events occurred in 91.7%. The most frequent grade ≥ 3 events were lymphopenia (66.7%), leukopenia (62.5%), neutropenia (62.5%), thrombocytopenia (58.3%), anemia (33.3%), and pneumonia (16.7%). Serious adverse events occurred in 10 patients (41.7%) and were considered related to treatment in 6 (25.0%). Infusion-related reactions occurred in 8 patients (33.3%), leading to interruption of IMM01 in 5, with no discontinuations. One patient (4.2%) died from a treatment-emergent adverse event related to study treatment, and the 30-day and 60-day mortality rates were 0 and 4.17%, respectively. Grade ≥ 3 anemia occurred in 33.3% of patients, with a single case of grade 3 hemolysis and no permanent discontinuations due to anemia or hemolysis.

The authors urge caution in interpreting these results. Recent phase 3 failures of magrolimab-based CD47-directed therapy in acute myeloid leukemia—the ENHANCE-2 and ENHANCE-3 trials—have prompted re-evaluation of the CD47–SIRPα axis, and mechanistic work by Chung and colleagues suggests that in human macrophages, anti-CD47 activity is primarily FcγR-mediated rather than due to CD47 blockade, with sialylated CD43 acting as an independent phagocytosis barrier. The study team acknowledges its own limitations: the single-arm design without a control group, the small sample of 24 patients, limited follow-up, and mutation heterogeneity all restrict generalizability. An in vitro assay using bone marrow cells from three enrolled patients showed that azacitidine pretreatment enhanced IMM01-mediated phagocytosis, but this mechanistic signal remains preliminary.

Subject of Research: A phase II trial of timdarpacept plus azacitidine as first-line therapy for higher-risk chronic myelomonocytic leukemia

Article Title: Timdarpacept (IMM01) combined with azacitidine (AZA) as first-line therapy in adults with higher risk chronic myelomonocytic leukemia (CMML): results from a prospective, multicenter, phase II study

Article References: Ma, L., Yang, W., Gao, S., Li, J., Yin, Q., Zhao, X., Yan, X., Tu, S., Li, F., Yang, H., Hu, R., Wang, S., Dong, Z., Zhao, X., Lu, G., Wu, Z., Tian, W., Tong, H., & Xiao, Z. (2026). Timdarpacept (IMM01) combined with azacitidine (AZA) as first-line therapy in adults with higher risk chronic myelomonocytic leukemia (CMML): results from a prospective, multicenter, phase II study. Experimental Hematology & Oncology, 15(1), Article 115. https://doi.org/10.1186/s40164-026-00851-8

Image Credits: AI Generated

DOI: 10.1186/s40164-026-00851-8

Keywords: timdarpacept, IMM01, azacitidine, chronic myelomonocytic leukemia, CD47, SIRPα, phase II trial, hypomethylating agents, macrophage phagocytosis, hematologic toxicity, overall response rate, clinical trial