Infusion of Ex Vivo Expanded Allogeneic Cord Blood-Derived Natural Killer Cells in Combination with Autologous Stem Cell Transplantation for Multiple Myeloma: Results of a Phase I Study

Background: Multiple myeloma (MM) is an incurable disease thought to be characterized by immune dysregulation and exhaustion, whereby proliferation of malignant plasma cells is not checked by the native immune system.  Long term remissions in some patients after allogeneic stem cell transplant (SCT) suggest a graft versus myeloma effect; however the treatment-related toxicity limits the widespread use of this modality.  Allogeneic natural killer (NK) cells are active in various hematologic malignancies and may have a role against MM, without concomitant graft versus host disease (GVHD). Umbilical cord blood is a potential source for allogeneic NK cells and ex vivo expanded umbilical cord blood-derived NK (CB-NK) cells demonstrate activity comparable to that of peripheral blood-derived NK cells.  We describe here the results of a phase I, first-in-human study of ex vivo expanded allogeneic CB-NK cells in conjunction with high dose chemotherapy and autologous SCT. 

Methods: Patients with symptomatic MM who were appropriate candidates for high dose chemotherapy and autologous SCT were eligible.  CB units with at least 4/6 match at HLA-A, B and DR were chosen for each patient.  When possible, CB units with potential NK alloreactivity were prioritized.  On day (-19) CB units were thawed and mononuclear cells (MNCs) were isolated by ficoll density gradient centrifugation.  MNCs were cultured in a gas permeable bioreactor with irradiated (100 Gy) K562-based aAPCs expressing membrane bound IL-21 “Clone 9.mbIL21” (2:1 feeder cell:MNC ratio) and IL-2 (100 IU/mL) .  On day 7, cells were CD3-depleted via immunomagnetic depletion and remaining cells were re-stimulated with aAPC feeder cells and cultured for an additional 7 days.  NK cell purity was determined after 14 days of culture.  Due to pre-clinical data demonstrating synergy between lenalidomide and NK cells, patients received lenalidomide (10 mg orally daily) from days (-8) to day (-2).  Melphalan 200 mg/ m² was given intravenously on day (-7).  Freshly expanded CB-NK cells were infused on day (-5).  Autologous peripheral blood progenitor cells (PBPC) were infused on day (0). 

Results: 12 patients have been enrolled thus far with 3 patients each on the following CB-NK cell dose levels: 5 e6 NK cells/kg, 1 e7 NK cells/kg, 5 e7 NK cells/kg and 1 e8 NK cells/kg.  11/12 patients had at least 1 high-risk feature of progressed/relapsed disease (n=7), high-risk cytogenetics (n=2) or International Staging System III disease (n=3).  Successful NK expansion to target dose was achieved in all patients with median NK purity (CD56+/CD16+/CD3-) of 98.9% (96.8-99.7).  Expanded cells demonstrated cytotoxicity against classic K562 and MM cell line targets. There were no infusional toxicities and no occurrence of GVHD.  One patient (1 e8 NK cells/kg) failed to engraft due to a poor PBPC graft quality; this patient was rescued with a back-up autologous PBPC graft.  There have been no other significant adverse events and no second primary malignancies.  11/12 patients are evaluable beyond day 100.  Best response has been 8/12 nCR or better, 2/12 VGPR and 1/12 PR.  4/12 patients have progressed at a median of 330 days.  By DNA microsatellite chimerism analysis, donor CB-NK cells were detected in 2 patients at the 1 e7 NK cells/kg dose and all 3 patients in the 1 e8 NK cells/kg dose, for at least 5 days after infusion.  By a more sensitive flow cytometric chimerism assay using HLA class I-specific antibodies for donor or recipient, donor CB-NK cells were detected in 3 evaluable patients at doses of 1 e7 NK cells/kg, 5 e7 NK cells/kg and 1 e8 NK cells/kg for at least 12 days after infusion.   Further analysis of these cells indicated persistence of an activated phenotype (NKG2D+/NKp30+) in vivo. 

Conclusion: CB-NK cells can be activated and expanded to clinical scale.  This is the first clinical study of CB-NK cells for MM.  When infused in the setting of myeloablative chemotherapy, up to 1 e8 allogeneic CB-NK cells/kg are well tolerated with no infusional toxicities or GVHD.  These cells can persist for at least 12 days in vivo and demonstrate an active phenotype.  Though clinical data are early, responses are encouraging in this high-risk patient population. Further updated data will be presented at the annual meeting.