IL-2 Stimulated Treg Inhibit in Vitro Expansion of Haploidentical Natural Killer (NK) Cells, Which Is Partially Overcome with An IL-2-Diphtheria Toxin Fusion Protein In Vivo,

Abstract 3611

Adoptive transfer of haploidentical natural killer (NK) cells can induce remissions in patients with refractory myeloid leukemia (AML). However, NK cells do not expand and persist in all patients despite lymphodepleting chemotherapy. In trials of adoptive NK cell therapy in solid tumors or lymphoma, host regulatory T cells (Treg) often expand in response to IL-2 given to stimulate donor NK cell expansion. Although murine studies report that Tregs inhibit NK cells, the influence of human Treg on NK cell proliferation and function is not well characterized. We studied the effect of allogeneic Tregs that were derived from human umbilical cord blood (UCB) as described by our group. Resting CFSE labelled NK cells or Teff were purified from healthy donors, and mixed with UCB Treg at various ratios. Unstimulated NK cells did not proliferate and thus IL-2 or IL-15 were added to the media at concentrations of 0.1, 0.25 and 0.5 ng/ml. In the absence of Treg, both cytokines induced equal NK cell proliferation at 5 days as measured by CFSE dilution in a concentration dependent manner. CFSE dilution was inhibited by Treg at a 1:1 ratio, especially at low cytokine concentrations. There were marked differences between the two cytokine conditions. Following IL-15 induced stimulation, the reduction in NK cell proliferation by Treg ranged from 1–35% (at different concentrations tested), whereas the inhibition of IL-2 stimulated NK cell proliferation ranged from 65–85%. Treg inhibition of NK cell proliferation could be measured at ratios as low as 1:8 in the presence of IL-2, but not IL-15. This inhibitory effect was partially explained by competition from CD25⁺ Tregs for IL-2. We measured Treg utilization of IL-2 by incubating NK cells with or without Treg in 0.5 ng/ml IL-2 for 4 days. The level of IL-2 with NK cells alone was 40 pg/ml vs. 17 pg/ml with Treg (compared to 330 pg/ml in IL-2-supplemented media without cells). Based on this data, we have incorporated host Treg depletion to enhance NK expansion after adoptive transfer to treat patients with refractory AML. As murine data from Blazar's group shows that CTL therapy is enhanced by Treg depletion, we added one dose of denileukin diftitox (ONTAK^®, Eisai Inc) at 12 mg/kg to our lymphodepleting preparative regimen of fludarabine 25 mg/m² × 5 days, cyclophosphamide 60 mg/kg × 2 days for 12 AML patients. Haploidentical NK cells (CD3- and CD19-depleted PBMCs and overnight activated with IL-2 1000 U/ml) were infused on Day 0, followed by 6 doses subcutaneous IL-2 (9 million units) given every other day to promote in vivo NK cell expansion. Eleven of 12 patients were evaluable, having received at least 4 of 6 planned doses of IL-2. Blood and marrow were collected 7 and 14 days after infusion to assess NK cell and Treg expansion, as well as leukemia clearance. Of the 10 patients with interpretable day 7 chimerism data, 9 had detectable donor DNA (median 68% donor DNA). At day 14, 4 of the 12 patients (33%) had successfully expanded NK cells in vivo, with absolute donor derived NK cell counts of 480, 530, 1470 and 12390 cells/μL blood, improving on our previous 10% rate of in vivo NK cell expansion which was observed with the same regimen, without Treg depletion. In the 4 patients who expanded NK cells in vivo, there were no detectable Treg (defined as a CD25⁺CD4⁺FoxP3⁺ lymphocyte population) at either day 7 or day 14. In contrast, the presence of a bona fide Treg population at either day 7 [range 9.5–53%] or day 14 [27–71%] correlated with a lack of in vivo NK cell expansion at day 14. Clinically, 8 of the 11 evaluable subjects cleared leukemia (72%), 7 of whom recovered neutrophils (63% CRp) and 6 of whom went on to best donor transplant (45%). In summary, we demonstrate in vitro and in vivo suppression of NK cell proliferation by IL-2 stimulated Treg. This effect is not seen in vitro with IL-15. We have shown that the absence of host Treg correlates with in vivo NK cells expansion. Although an increased rate of donor NK expansion was observed with a single dose of denileukin diftitox, it did not completely overcome the IL-2 induced host Treg expansion. Future trials testing additional doses of denileukin difitox or other methods of Treg depletion, as well as the use of IL-15 are planned.