Introduction Allogeneic stem cell transplantation has demonstrated the ability to prevent leukemic relapse via an immune-mediated graft-vs.-leukemia effect. Natural killer (NK) cells have been shown to comprise a significant component of this anti-leukemia effect and have been reported to enhance engraftment and reduce graft-vs.-host disease. Ex-vivo expansion of peripheral blood-derived NK cells has been demonstrated. Cord blood (CB) is a promising alternate source of NK cells with enhanced cytokine/antigen-responsiveness, proliferation and increasing availability. We studied expansion conditions for the generation of CB NK cells for clinical use, and evaluated them in vivo in NOD-SCID/IL2Rgnull mice engrafted with human leukemia.

Methods Two expansion protocols were compared:

  1. CD56+ NK cells were isolated and cultured with irradiated CD56 feeder cells in RPMI-1640 supplemented with 100ng/mL IL-2 for 21 days.

  2. CB mononuclear cells were depleted of CD3+ cells prior to culture.

In vivo NK cytolytic activity was measured via chromium release assays on leukemia cell lines (K562, Nalm6) and fresh patient leukemia blasts. Functionality was confirmed in vivo as well by injecting 106 K562 cells or 2.5 ×104 Nalm6 cells into NOD-SCID/IL2Rgnull mice. 2 ×106 CB-expanded NK cells were administered on day +1, and an additional 5×106 CB NK cells were administered on day +7. NK killing of leukemic targets was then confirmed by flow cytometry.

Results Feeder cells mediated >30-fold expansion of CB NK cells, generating a nearly pure population that contained 97% CD3CD56+ cells. Expanded CB NK could lyse >90% of K562 targets at an E:T ratio of 10:1, as well as >40% of patient leukemia blast. In CD3 depleted cultures, similar expansions were observed, however only 55% of these cells were CD3CD56+ with the remainder being 10% CD3+CD56+ NK/T and 35% CD3+ T lymphocytes. In the NOD-SCID~/IL2Rgnull model, ex vivo expanded CB NK cells demonstrated the ability to reduce the leukemia burden of both AML (K562) and ALL (Nalm6) cells >50%.

Conclusions We have demonstrated the feasibility of expanding CB NK ex vivo to clinically-relevant doses with minimal manipulation. These expanded CB NK cells demonstrated cytotoxic activity in vitro and in vivo against a variety of human leukemic cell lines and patient leukemia blasts. These results provide rationale for immunotherapy of leukemia with CB-derived NK cells.

Author notes

Disclosure: No relevant conflicts of interest to declare.