The Bispecific Innate Cell Engager AFM28 Can Leverage AML Patient’s NK Cells in Addition to Allogeneic NK Cells, Enabling Elimination of CD123+ Leukemic Stem and Progenitor Cells in AML and MDS

Introduction: Acute myeloid leukemia (AML) and high-risk myelodysplastic neoplasms (HR-MDS) remain challenging hematological malignancies with poor long-term outcomes. Novel strategies targeting leukemic stem and progenitor cells (LSPCs) are crucial to improve durable remissions. In that regard, CD123 expressed on LSPCs and leukemic blasts has emerged as a promising therapeutic target. We have demonstrated the preclinical efficacy of the bispecific CD123/CD16A innate cell engager AFM28 against CD123+ leukemic cells mediated by natural killer (NK) cells including its favorable safety outcome in cynomolgus monkey studies. Here, we investigated whether AFM28 alone could be preclinically effective by leveraging AML patient-derived NK cells to target autologous leukemic cells in addition to its potential in combination with allogeneic NK cell therapy. Moreover, we explored the preclinical effects of AFM28 on hematopoietic progenitors and LSPC derived from AML and HR-MDS patients.

Methods: To assess the activity of AFM28 against leukemic blasts and LSPC, primary bone marrow (BM)-derived mononuclear cells (MNCs) of AML (n=5) and HR-MDS (n=5) patients were co-cultured with allogeneic NK cells from healthy donors in the presence of AFM28 for 24 h, followed by flow cytometric analysis of CD123+ leukemic cells including LSPCs defined as CD123+CD34+CD38−CD117+ cells. Colony forming unit (CFU) assays were performed using CD34+ MNCs derived from AML patients (n=5), HR-MDS patients (n=5) and healthy donors (n=5) after exposure to allogeneic NK cells and AFM28. In ex vivo assays, freshly drawn blood of newly diagnosed AML patients (n=7) was treated with AFM28 for 24 h with or without additional allogeneic NK cells.

Results: In an autologous setting, AFM28 demonstrated activity against primary leukemic blasts in ex vivo cultures of fresh whole blood derived from untreated first diagnosis AML patients, indicating that AFM28 has the potential to arm AML patient’s endogenous NK cells for anti-leukemic activity. Responder samples (n=3) showed a mean lysis of 76% (range 59-90%), while non-responders (n=4) exhibited a mean lysis of 7% (range 0-22%). The addition of healthy allogeneic NK cells could convert non-responders to responders and enhanced the overall anti-leukemic activity to 89% (range 77-100%) in previous responders and 66% (range 18-99%) in previous non-responders. Moreover, AFM28 induced the elimination of 98% of LSPCs (range 95–99%) in cultures using BM-MNCs of AML and HR-MDS patients supplemented with allogeneic NK cells. Depletion of LSPCs was corroborated by the concentration-dependent reduction in leukemic colony formation of AML (57-75% reduction) and HR-MDS BM-MNCs samples (57-64%), whereas colony formation of healthy BM-MNCs samples was only modestly affected (6-39%). Hence, AFM28 demonstrated specificity for leukemic cells while largely sparing healthy hematopoietic progenitors.

Conclusion: This study highlights the potent efficacy of AFM28 in targeting CD123+ leukemic cells including LSPCs of AML and HR-MDS patients. Our results demonstrate that AFM28 monotherapy has the potential to mediate anti-leukemic activity by leveraging the endogenous innate immunity in some AML patients. The anti-leukemic activity of AFM28 may be further increased in combination with allogeneic NK cell immunotherapy to improve responses in patients with low numbers and/or hypofunctional NK cells.