Treatment options for refractory AML are usually ineffective. Previously, we tested adoptive transfer of haploidentical peripheral blood (PB) derived NK cells without transplantation and demonstrated correlation between in vivo NK cell expansion and those who achieved a complete remission. This therapy is limited by:

  1. the inability to expand NK cells in most patients,

  2. prolonged neutropenia in some patients and

  3. inconsistent efficacy.

UCB, in contrast to adult PB, is rich in NK precursors with CD34+/CD7−, CD34+/CD7+ and CD34−/CD7+ phenotypes. We hypothesized that UCB-derived NK cells may show better in vivo expansion than adult derived NK cells after cytoreduction. Therefore, we tested our triple UCBT strategy in patients with refractory relapsed AML who were <45 years old, without active infection and eligible for myeloablative conditioning. The UCB NK product (unit 1) was CD3 depleted (using a CliniMacs system) and activated with IL-2 (1000U/ml for 16–20 hours). The UCB-derived NK cells (matched at 3 HLA loci and KIR-ligand mismatched when possible) were infused on day -12 after conditioning with cyclophosphamide 120 mg/kg, fludarabine 125 mg/m2 and TBI 1320 cGy on days -19 to -13. Subcutaneous IL-2 (10 MU) was was given on days -12, -10, -8, -6, -4 and -2 to facilitate in vivo NK cell expansion. On day 0, two UCB units (≥4/6 match) were transplanted for hematopoietic rescue and followed by mycophenolate mofetil and cyclosporine for GVHD prophylaxis. Compared to pre-treatment levels, endogenous IL-15 was markedly increased after the preparative regimen at the time of the NK UCB unit. The NK UCB units contained both precursor and mature NK cells. Three product samples were cultured for 28 days in limiting dilution on a murine stromal feeder, demonstrating cloning frequencies of 1:5, 1:9 and 1:12 infused UCB cells giving rise to NK progeny. Two of the 3 patients had partial chimerism derived from the NK product on day -1. Unexpectedly, these same two patients demonstrated prompt neutrophil engraftment on days 3 and 7 after hematopoietic stem cell rescue. In both instances, chimerism was achieved from the NK product. Of the non-NK and non-T cells in the NK UCB units from these 2 patients there were 9.6% and 5.3% CD34+/CD7− cells. In the third patient, the NK UCB unit had only 2% CD34+/CD7− cells and they did not contribute to neutrophil engraftment which occurred on day 36 after UCBT. Of note, 3/3 tolerated the NK infusion without toxicity and were leukemia-free at the time of engraftment. Two remain alive (one died of TRM) with one relapse before day 100. These data suggest UCB NK cells may be administered safely and, despite CD3 depletion and IL-2 activation (ex vivo and in vivo), provide long term engraftment potential that may dominate over unmanipulated UCB infused subsequently. In summary, UCB is a rich source of NK precursors capable of in vivo expansion which are potentially better suited than adult NK cells for use in treatment of patients with refractory AML.

Disclosure: No relevant conflicts of interest to declare.

Author notes


Corresponding author