Successful Remission of Poor Prognosis AML after Adoptive Transfer and In Vivo Expansion of Human Haploidentical NK Cells.

We previously demonstrated that although autologous NK cell therapy after hematopoietic cell transplantation (HCT) is safe, it does not provide an anti-tumor effect. We hypothesize that the lack of NK cell inhibitory receptor (KIR) mismatching with autologous tumor cells limits anti-tumor efficacy. Allogeneic NK cell infusions may overcome this restriction. Here we tested haploidentical, related donor NK cell infusions in a non-transplant setting to establish safety with a goal of in vivo NK cell expansion. A low intensity outpatient immune suppressive regimen of fludarabine (Flu) alone was tested in 8 renal cell carcinoma patients. A higher intensity inpatient regimen of high-dose cyclophosphamide and Flu (Hi-Cy/Flu) was tested in 12 patients with poor prognosis AML, all with active leukemia and most who failed 1 or more standard theries. NK cells were enriched from donor apheresis products using CD3 depletion by the CliniMacs system. All patients received subcutaneous IL-2 after infusions. Patients who received the lower intensity Flu regimen showed transient persistence of donor cells, but no in vivo expansion (<0.1%) using a PCR based chimerism assay of unique MHC class I alleles in the donor but not present in the recipient. There were no objective responses in renal cell carcinoma patients. In marked contrast, 6 of 7 AML patients who were evaluable for engraftment receiving Hi-Cy/Flu showed engraftment of donor cells, some for > 28 days. Three of 6 patients had peak engraftment levels of 10%, 100% and 100% of PBMC in the first 2 weeks after the haploidentical infusion. The remaining three patients had lower levels of approximately 1%. Five of 12 patients have achieved a remission with this therapy. Compared to patients who failed to clear their leukemia, the remission patients had significantly higher proportions of circulating NK cells (35±8% vs. 1.5±0.4%, p=.001) and cytotoxicity against K562 targets (50±10% vs. 10±5% at E:T 20:1, p=0.01) suggesting that in vivo expansion was required to achieve efficacy. We next tested whether endogenous cytokines play a role in the in vivo expansion. Based on its established role in NK cell differentiation and on recent mouse studies, IL-15 was a likely candidate to drive homeostatic NK cell expansion. Flu patients showed only a slight increase in IL-15 after several days compared to pre-treatment levels. In marked contrast, AML patients had significant increases in plasma IL-15 after Hi-Cy/Flu therapy, which appeared before haploidentical cell infusions and was sustained for several weeks. There was an inverse correlation between the absolute lymphocyte count and the IL-15 concentration (r = −.62, p-value < .0001). We conclude that Hi-Cy/Flu therapy allows for expansion of allogeneic NK cells in vivo in part through increased endogenous IL-15 concentrations, which can induce remissions in poor prognosis AML patients. For the first time, these findings suggest that haploidentical NK cells can persist and expand in vivo and may have a role in the treatment of AML used alone or as an adjunct to HCT. Future studies will address NK cell donor-recipient mismatch at KIR ligands, patient selection and other variables that impact on response.