Introduction

Elderly acute myeloid leukemia (AML) patients have a poor prognosis due to high relapse rates following standard therapy. Natural Killer (NK) cell alloreactivity has found to control relapse in AML in the HLA-mismatched haploidentical allogeneic stem cell transplantation (allo-SCT) setting. Moreover, allogeneic NK cell infusions can induce complete remission (CR) inpatients with advanced AML. As a consequence, adoptive NK cell transfer may be a promising treatment for elderly AML patients, who are not eligible for allo-SCT. Most clinical studies exploited NK cell products enriched from leukapheresis of haploidentical donors containing low numbers of T cells that could have contributed to the observed therapeutic effects and potentially induced graft-versus-host disease (GVHD). Therefore, we have developed a GMP-compliant culture system for the generation of large batches of NK cells from umbilical cord blood (UCB)-derived CD34+ progenitor cells, without T cell contamination. Here, we report results of a phase I dose escalation study (Dutch Trial Register nr. NTR2818) to evaluate the feasibility, safety and toxicity of allogeneic UCB-NK cell infusion following an immunosuppressive preparative regimen in elderly AML patients. Secondary endpoints were NK cell lifespan and the effects on minimal residual disease (MRD).

Methods

Elderly AML patients not eligible for allo-SCT, and in morphologic CR after standard therapy, were given preparative chemotherapy consisting of Cyclophosphamide (Cy;900 mg/m2/day) and Fludarabine (Flu;30 mg/m2/day) on days -6 to -2. At day 0, UCB-NK cells at a dose of 3, 10 or up to 30x106/kg body weight were infused without IL-2 treatment to study if in vivo expansion could be obtained without IL-2 support. Patients were assessed for toxicity and GVHD. Donor chimerism was measured by Q-PCR for discriminating DNA polymorphisms. NK cell expansion and phenotype were analyzed by flow cytometry. MRD was evaluated by flow cytometry and molecular techniques.

Results

Twelve AML patients (68-76 years) have been included, all in morphologic CR after 2 to 3 standard chemotherapy courses (n=6), or 1 standard chemotherapy course followed by subsequent treatment with hypomethylating agents (azacitidine or decitabine) (n=6). Patients were treated with Cy/Flu and an escalating dose of partially HLA-matched UCB-NK cells. Four patients had good/intermediate risk, 4 poor risk and 4 very poor risk AML. To date, 9 patients received NK cell products containing a median of 74% highly activated CD56+ NK cells, with <1x104/kg CD3+ T cells and <3x105/kg CD19+ B cells. Remaining non-NK cells were CD14+ and/or CD15+ monocytic and myelocytic cells. Follow up did not show GVHD or toxicity attributed to the NK cells. As expected, preparative Cy/Flu induced a neutropenic period of 20 ± 16 days, but no severe infections were seen. A temporary repopulation and persistence of UCB-NK cells could be detected in peripheral blood between days 1 and 8 post-infusion, which was associated with increased IL-15 plasma levels observed in most patients. Interestingly, donor chimerism increased with higher doses of infused UCB-NK cells, and donor chimerism up to 3.5% was found in bone marrow (BM) at day 7/8. Further UCB-NK cell maturation in vivo was observed by acquisition of CD16 and KIRs, while expression of activating receptors was sustained. Of the 9 treated patients so far, 5 (56%) are still in CR after 43, 35, 31, 5 and 4 months, whereas 4 patients relapsed after 5, 6 (2 pts) and 15 months. Despite morphologic CR during azacitidine treatment, residual disease of 6-7% with a leukemia-associated phenotype could be detected by flow cytometry before NK cell infusion in BM of two patients. In both patients MRD was reduced to less than 0.05% at 90 days after UCB-NK cell therapy following Flu/Cy conditioning.

Conclusion

These results show that GMP-compliant UCB-NK cell products containing up to 30x106 NK cells/kg body weight can be safely infused in non-transplant eligible AML patients following immunosuppressive chemotherapy. After infusion, UCB-NK cells repopulate, mature and migrate to BM without supporting IL-2 infusion. Since we observed reduction in MRD in patients on treatment with hypomethylating agents, this UCB-NK cell therapy may induce or sustain CR in elderly AML patients, and could serve as an alternative consolidation therapy for patients with refractory AML or provide bridge to allo-SCT.

Disclosures

Spanholtz:Glycostem Therapeutics: Employment. Tordoir:Glycostem Therapeutics: Employment. Bohme:Glycostem Therapeutics: Employment. Kok:Glycostem Therapeutics: Employment.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution