Killer Immunoglobulin-Like Receptor Ligand (KIR-Lig) Mismatched Natural Killer (NK) Cell Transfusions for Multiple Myeloma (MM).

Abstract

Remarkable observations in haplo-identical, T-cell-depleted allogeneic transplantation (Tx) for AML have focused attention on the anti-tumor effects mediated by KIR-lig mismatched NK cells. In this study we evaluated whether haplo-identical KIR-lig-mismatched NK cells transfused after immunosuppression and tumor reduction with high dose chemotherapy and followed by a delayed auto-Tx can improve outcome in high risk MM patients. All 4 patients enrolled had poor prognosis MM (cytogenetically abnormal, high CKS1-B expression) relapsing after tandem (n=3) or single auto-Tx (n=1). The conditioning regimen comprised fludarabine (immunosuppression, days −5 to −2), and melphalan (tumor reduction, day -1). Patients who lacked HLA known to inhibit donor NK cells (HLA-Bw4, or -C group I or II) received IL2 activated CD3 depleted apheresis products from a haplo-identical family donor (days 0 and 2). IL2 was given (days 1 to 11) to support NK cell survival in vivo. An autograft, delayed until day 14 to allow time for NK cells to kill MM, provided hematopoetic and immunologic reconstitution. Absolute CD3−56+ NK cell doses were 37,100,39, and 95x106/kg, while CD3 doses were limited to ≤ 6.2x104/kg. Genotyping and phenotyping confirmed the presence of donor NK cells expressing the relevant KIRs. The anti-tumor potency of NK cells was confirmed by their ability to kill K562, patient MM (when available, n=2) and U266 MM cells but not patient non-MM cells, at low effector to target ratios ≤ 10:1. Only one severe adverse event was seen, a transient TRALI-like event caused by a red cell lysis step which was subsequently omitted from the manufacturing procedure. No patients developed GvHD. Real time PCR for HLA or STR analysis revealed donor cells in the patient's peripheral blood on day 7 (36, 88, 27, 90% for patients 1–4) that were no longer detectable by day 14. We identified proliferating lymphocytes of patient origin which eliminated the donor cells prior to auto-Tx (specific anti-donor reactivity was confirmed by mixed lymphocyte analysis). The treatment regimen resulted in a near complete response (nCR) or CR in all patients, with remarkable resolution (by FDG-PET) of ≥42 focal lesions in 3 patients. All 4 patients suffered an early relapse (day 90,153,42,40), but 3 patients subsequently achieved nCR/CR with salvage chemotherapy VTD(A). This is the first clinical application of KIR-lig-mismatched haplo-identical NK cell transfusions in autologous Tx and the first such trial in MM. Interestingly, after therapy, 2 patients with previously refractory disease had become chemosensitive. We demonstrated that haploidentical donor NK cells kill patient MM cells. However, NK cell doses transfused may not be sufficient to destroy a large MM tumor burden in vivo, which is consistent with findings observed in AML after haplo-identical Tx. Therapeutic efficacy may be enhanced by treating patients with a lower tumor burden and/or providing additional immunosuppression to abrogate the anti-donor response. Lastly, we have preliminary data that it is possible to specifically expand and activate NK cells with enhanced anti-MM effects using K562-membrane bound IL15-41BB ligand transfectants.