Umbilical cord blood CD34+ hematopoietic stem cell-derived natural killer cells in Acute Myeloid Leukemia – a phase I clinical study

Abstract Background: The use of fresh umbilical cord blood (UCB) CD34+ hematopoietic stem cell (HSC)-derived allogeneic natural killer (NK) cells has shown favorable safety in a phase I clinical trial in patients with acute myeloid leukemia (AML). Methods: In a multicenter, open-label, phase I study (NCT04632316) we evaluated the safety and efficacy of cryopreserved UCB CD34+ HSC-derived NK cells, inaleucel (oNKord®), after lymphodepleting conditioning regimen of cyclophosphamide/fludarabine (Cy/Flu) (for cohorts A1, A2, A3, A5), or azacitidine/venetoclax (Aza/Ven) (A4) followed by 1, 2 or 3 NK cell infusions in AML patients. Patients had to be in morphologic complete remission (CR) (including CR with incomplete hematologic recovery [CRi]) with measurable residual disease (MRD), who were not proceeding to allogeneic hematopoietic cell transplantation (alloHCT). MRD was assessed by flow-cytometry and ultrahigh-sensitivity NGS. NK cell persistence was assessed by ultra-high sensitivity molecular chimerism analysis. Results: The intention-to-treat population included 16 AML patients (10 male, 6 female) with a median age of 74.5 years (range 45-82). Patients were in first or second CR after intensive (n=6) or non-intensive (n=10) chemotherapy. The distribution of ELN risk groups varied across cohorts, with a notable enrichment of favorable-risk patients in Cohort A3 (3/4) and A5 (2/3), whereas intermediate and adverse risk groups were more evenly represented in Cohorts A1, A2, and A4. Two patients had a TP53 mutation (A1 and A4) and one patient in group A3 had t(8;21)(q22;q22.1). Three cohorts (A1-3) of 3 patients each received escalating doses of NK cells (3 × 325 – 1,000 × 106 viable NK cells) on D0, D4 and D8 after Cy/Flu conditioning. Additionally, the fourth cohort (A4) received the same dose as the third cohort (A3), but followed an adjusted conditioning strategy with Aza/Ven, while the fifth cohort (A5) received the same cumulative total dose of NK cells as A3 but as a single infusion on D0. The NK cell therapy was well-tolerated and safe, with no adverse events related to the NK cells reported in A1, A2, A4 and A5, and one grade 1 cytokine-release syndrome observed in a patient of A3, which was attributable to pneumonia. This correlated well with the detection of low levels of cytokine IL-6 in patient sera. Unlike previous observations, in this cohort, patient sera analysis did not show any increase in IL-15 levels after cyclophosphamide (300 mg/m2/day at day -5 to -3) and fludarabine (30 mg/m2/day at day -5 to -3) conditioning, and IL-2 levels were overall low. Despite the lack of intrinsic or extrinsic cytokine support, NK cells were detectable for up to 14 days post infusion. MRD dynamics assessed by multiparameter flow cytometry (MFC) showed that patients with baseline MRD positivity had a median burden of 0.26% of BM CD45⁺ cells (range up to 2.3%). Notably, MRD negativity was achieved in 41% (5/12) of Cy/Flu-preconditioned patients whereas none (0/3) of the Aza/Ven-preconditioned patients reached MRD-negativity status by Day 30 post-oNKord infusion. Those who maintained MRD negativity (n=3) demonstrated sustained clinical benefit and completed the study, highlighting a strong association between early MRD clearance and improved outcomes. Furthermore, to assess early treatment efficacy, we monitored variant allele frequencies (VAFs) of all known mutations using NGS-MRD analysis. Among 16 patients, 7 (44%) exhibited >50% reduction in VAFs for all detected variants between baseline and 3 months post-treatment. In 5 additional patients (31%), at least one mutation showed >50% reduction in VAF, while other mutations persisted. The remaining 4 patients (25%) demonstrated persistent VAFs for all detected mutations, including IDH2 and TP53. The one-year survival was 50% for all patients (n=16) and 61.5% for Cy/Flu preconditioned patients. After a median follow-up of 33.4 months, including the patients that also participated in the follow up ReKORD study NCT05290662, the median overall survival of the total cohort was 13.1 months since study entry (n=15) and 14.9 months for Cy/Flu preconditioned patients (n=13).Conclusions: One to three infusions of cryopreserved UCB CD34+ HSC-derived NK cells, inaleucel, in MRD positive AML patients were determined safe and well tolerated. Based on these promising signs of safety and efficacy dose expansion with Cy/Flu pre-conditioning is planned for the near future.