Treatment of patients with advanced cancer with the natural killer cell line NK-92

Abstract

Background aims

Natural killer (NK) cells, either naive or genetically engineered, are increasingly considered for cellular therapy of patients with malignancies. When using NK cells from peripheral blood, the number of expanded NK cells can be highly variable and the need for NK cell enrichment can make the process expensive. The NK-92 cell line (CD56+/CD3−) that was isolated from a patient with lymphoma has predictable high cytotoxic activity and can be expanded under good manufacturing practice conditions in recombinant interleukin-2.

Methods

Fifteen patients (age, 9–71 years) with advanced, treatment-resistant malignancies, either solid tumors/sarcomas (n = 13) or leukemia/lymphoma (n = 2), received two infusions of NK-92 cells, given 48 h apart. Three cohorts of patients were treated with escalating doses of NK-92 cells (n = 7 at 1 × 10⁹, n = 6 at 3 × 10⁹ and n = 2 at 1 × 10¹⁰ cells/m²).

Results

No infusion-related or long-term side effects were observed. The dose of 10¹⁰ cells/m² was considered the maximum expandable cell dose with the use of an established culture bag system. Three fourths of patients with lung cancer had some anti-tumor response. Only one patient of seven had development of human leukocyte antigen antibodies. The persistence of NK-92 cells (male origin) in the circulation was confirmed by Y chromosome–specific polymerase chain reaction in two female patients.

Conclusions

Infusions of NK-92 cells up to 10¹⁰ cells/m² were well tolerated. Despite the allogeneic nature of NK-92, development of human leukocyte antigen antibodies in these patients with cancer appears to be rare. The cells can persist in the recipient's circulation for at least 48 h. Some encouraging responses were seen in patients with advanced lung cancer.

Introduction

Chemotherapy is not curative for most cancers because it only affects dividing cells, allowing the true cancer stem cells with slow or minimal turnover to remain unaffected. Cancer cells also develop resistance to drugs, which ultimately contributes to disease recurrence. Although new drug developments hold some promise, thus far the survival benefit for non-hematological cancers has been marginal, largely because the malignant clone rapidly adjusts to the pathway with which the drug interferes.

Cell-based therapies involving cytotoxic cells are believed to attack and kill malignant stem cells through alternative mechanisms that include DNA degradation through the perforin/granzyme pathway and release of immuno-stimulatory (ie, interferon [IFN]-γ) or cytotoxic cytokines (ie, tumor necrosis factor [TNF]-α) (1). These pathways are cytolytic to target cells independent of cell cycle status and cell divisions and are not cross-reactive with other anti-cancer treatments. The cells of the immune system—largely T lymphocytes and natural killer (NK) cells—are now recognized as being able to provide control of cancer cell growth. Patients in whom leukemia relapses after an allogeneic hematopoietic stem cell transplant (HSCT) can achieve a remission when donor lymphocytes are infused 2, 3. Furthermore, when comparing the incidence of relapse after syngeneic and allogeneic transplant, there is an approximately 30% higher recurrence rate after syngeneic transplant because of the identical genetic make-up of the transplanted cells that do not evoke a donor-derived cellular immune response (4).

More recently, NK cells have received some attention because they recognize tumor cells through a different pathway than T-lymphocytes. Generally, the activity of NK cells is blocked by killer cell immunoglobulin receptors (KIR) when they encounter cells with the cognate (“self”) major histocompatibility complex (MHC) pattern 5, 6. Conversely, they can become activated when they encounter non-self. This was first observed in the context of MHC haplotype-mismatched HSCT when T-lymphocytes were removed from the donor marrow 7, 8. This observation was recently confirmed in a large, retrospective analysis with patient/donor samples from the National Marrow Donor Program (9). Transplantation of donor/recipient KIR/ligand mismatches provided a survival advantage because of a lower relapse rate. In the non-transplant setting, infusion of allogeneic, CD3-depleted mononuclear cells enriched for CD56+ NK cells could induce remission in some 25% of patients with acute myeloid leukemia (10). Allogeneic MHC-mismatched NK cells can also be given to patients as part of a sibling HSCT (11) as well as early after an autologous HSCT without concern of graft-versus-host disease or marrow suppression (12), provided that T cells are removed.

Several laboratories are actively working on isolating and expanding clinical scale numbers of NK cells for treatment of patients with cancer (13). Since infusions of autologous NK cells, even with concomitant administration of cytokines such as interleukin (IL)-2, are clinically ineffective (14), the focus has shifted to allogeneic NK cells from a related or unrelated donor. Because the majority of peripheral lymphocytes obtained by apheresis are T-cells, they must be removed to prevent graft-versus-host disease in the recipient, which adds another step of manipulation and expenses to the expansion process (13). There is also substantial donor-dependent variability in terms of NK cell yield 12, 13, 15, 16, 17. For these reasons, a continuously growing NK cell line, providing predictable numbers of highly cytotoxic NK cells on expansion, is attractive for cellular therapy.

The activated NK cell line NK-92 is highly cytotoxic against a broad spectrum of malignant cells 18, 19, 20, 21, 22; the cells lack all currently known inhibitory KIR receptors, except for KIR2DL4, which is expressed at low levels, while expressing the full spectrum of activating receptors (22). The expansion kinetics of NK-92 have been delineated, and predictable numbers of activated NK-92 cells can be reproducibly obtained after culture in IL-2 (23). A clinical phase I trial with three infusions given every other day has been reported in patients with advanced renal cell cancer and melanoma (24), and preliminary results from this trial have also been presented (25). The infusions of NK-92 cells were well tolerated, and some of the patients with end-stage cancer showed some tumor responses.

In the present report, we show results from a phase I trial with NK-92 cells that were administered to a group of patients with predominantly end-stage solid tumors and with a different infusion schedule than previously reported (24). This study also tested for the presence of human leukocyte antigen (HLA) antibodies and persistence of NK-92 cells in the circulation of selected patients. Infusions of NK-92 cells were well tolerated, even after re-treatment at a later time, with some patients with advanced chemotherapy-resistant lung cancer having disease responses.