Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2015 Jan 1.
Published in final edited form as: Crit Rev Oncog. 2014;19(0):91–105. doi: 10.1615/critrevoncog.2014011142

NK Cells in the Tumor Microenvironment

Stine K Larsen 1,2, Yanhua Gao 3, Per H Basse 1,4,*
PMCID: PMC4062922  NIHMSID: NIHMS590354  PMID: 24941376

Abstract

The presence of natural killer (NK) cells in the tumor microenvironment correlates with outcome in a variety of cancers. However, the role of intratumoral NK cells is unclear. Preclinical studies have shown that, while NK cells efficiently kill circulating tumor cells of almost any origin, they seem to have very little effect against the same type of tumor cells when these have extravasated. The ability to kill extravasated tumor cells is, however, is dependent of the level of activation of the NK cells, as more recent published and unpublished studies, discussed below, have demonstrated that interleukin-2–activated NK cells are able to attack well-established solid tumors.

Keywords: Natural killer cells, infiltration, tumor microenvironment, prognosis, IL-2, virus, anti-tumor effect

I. INTRODUCTION

The discovery of natural killer (NK) cells and naturally occurring cytotoxicity against tumor cells1,2 immediately generated high expectations for the efficacy of immunotherapies based on these effector cells. An early finding in animal models indicated that NK cells participate in the defense against metastases. It was demonstrated that tumor cell clearance in the lungs, in large part, relied on NK cells and that the rate of the NK-cell–mediated clearance of tumor cells in the lungs inversely correlated with the number of experimental metastases that would develop in the lungs and most other organs following intravenous (i.v.) injection of tumor-cell suspensions.3–6 Because the portal vein drains venous blood from the gut, pancreas, and spleen, it would seem beneficial if NK cells were present also in the liver to protect it from tumor cells that are released to the portal circulation. In fact, an inverse correlation between the rate of tumor cell clearance from the liver and the number of liver metastases has been demonstrated,7,8 but more studies are needed to determine whether a similar inverse correlation between rate of clearance and number of metastases exists in other organs. Whether or not this turns out to be the case, an efficient NK-cell–mediated clearance of tumor cells in the lungs, through which all venous blood passes, is a logical way of protecting the host from metastasis. As early as 1990, Vaquer et al. found that a decrease in NK activity present in the blood from women with uterine cancer coincided with dissemination of the primary tumor.9

Numerous studies subsequently demonstrated that treatment of mice with NK activators, for example, Poly I:C (a TLR-3 agonist), substantially improved the mice's resistance to metastases.10–16 However, because tumor cells may be shed and become blood-borne at any time, it is not simple, from a clinical point of view, to take advantage of this observation. To this end, the benefit of boosting the NK activity of cancer patients at least in those periods where a heightened number of circulating tumors cells or a reduction of the patient's NK activity can be expected (e.g., during surgical removal of tumors) has long been suggested17–27 but has not yet been confirmed in clinical studies. However, more than a decade ago it was shown that preoperative NK-cell activity is a prognostic factor for distant metastasis following surgery for colon cancer.28

Disappointingly, most attempts to eliminate already established tumors and metastases by augmentation of host NK cell activity failed.29–31 Due to these disappointments, it has long been the paradigm that, while NK cells are important as a first line of defense against metastasis from circulating tumor cells, they do not play any major role in the defense against tumor cells that reside in the tissues, i.e., outside the blood vascular system.

In this mini-review, we discuss the conundrum that, although NK cells efficiently kills circulating tumor cells of almost any origin, they seem to have very little effect against the same type of tumor cells when these have extravasated. We will also discuss newer published and unpublished findings that challenge the current paradigm that NK cells are incapable of attacking already established solid tumors.

II. THE DENSITY OF NK CELLS IN SOLID TUMORS CORRELATES WITH PROGNOSIS

The role of NK cells in cancer may not, as previously thought, be restricted to only blood-borne tumor cells; newer findings implicate NK cells in the control of extravascular tumor growth as well, e.g., in the early phases of oncogenesis. In a study by Guerra et al., tumors developed faster in models of spontaneous leukemia and prostate cancer in animals depleted of NK cells compared to animals with normal NK cells activity.32 Although the presence of NK cells in the pre- and/or early malignant lesions was not investigated, fully developed tumors from NK-cell–deficient mice expressed ligands for the activating NK receptor NKG2D, whereas tumors from NK-competent mice did not express these ligands. This finding indicates that tumors developing in these animals (a process which, at least in the pancreas model, does not involve blood-borne tumor) had been edited by NK cells.

The prognostic significance of NK cells in patients with fully established colorectal carcinomas was first demonstrated by Coca et al.33 Patients with little and moderate NK infiltration had significantly shorter survival rates (overall, and disease free survival) than those with extensive infiltration. In addition, the density of tumor-infiltrating NK cells appeared to be a prognostic factor in the survival of patients with squamous cell lung cancer.34 However, in a recent study of 20 male patients who underwent surgery to remove a single cerebral metastasis from lung adenocarcinoma, there was no correlation between the degree of NK-cell infiltration within resected brain metastases and the period free of intracranial disease after surgery.35 Nevertheless, the list of cancer types in which a correlation between intratumoral NK-cell density and prognosis has been found, is steadily growing (Table 1). These and similar studies show that NK cells indeed are able to not only localize into extravascular, solid tumors but that they also add functionally to the host–tumor relationship. While it seems very likely that host NK cells continue to kill pre- and/or early-malignant cells in small developing tumors until the tumor is either eliminated or until it has been edited to a point where it no longer expresses sufficient NK ligands to elicit NK-cell–mediated cytotoxicity, the causality between NK-cell density in larger tumors and prognosis remains an open question. Based on the available information, we have tried to estimate the densities of intratumoral NK cells, which, in the studies listed in Table 1, have been associated with favorable prognoses. In several studies, 30 or more NK cells mm−2 of tumor tissue was arbitrarily considered a high density. Assuming that the volume fraction of the malignant cells in most tumors is at least 25%, a minimum of 1,000 tumor cells will be visible per square millimeter of tumor in sections of standard thickness (5–8 μm). Thus, the estimated effector-to-target (E:T) ratio in high-NK-density tumor tissue is, at best, 1:35. It is doubtful whether cytotoxicity or cytokine-secretion by NK cells present at such low densities has any significant impact on tumor growth. It is possible that the presence of a high number of NK cells in tumors indicates that the patient's NK cells are in a good shape overall and that they therefore are able to contribute to the host's anti-tumor immune responses, either by eliminating circulating tumor cells or by secreting the appropriate, perhaps Th1-stimulating, cytokines during cross talk with dendritic cells in the periphery and/or in the secondary lymphoid tissues.

TABLE 1.

Correlation between intratumoral NK-cell density and prognosis in various cancer types

Cancer type Reference
Colorectal carcinoma Coca et al. 1997 33
Hepatocellular carcinoma Taketomi et al. 1998 70
Gastric carcinoma Ishigami et al. 2000 71
Adenocarcinoma lung Takanami et al. 2001 72
Gastric carcinoma Takeuchi et al. 2001 73
Leukemia Lowdell et al. 2002 74
Squamous cell, lung Villegas et al. 2002 34
Renal cancer Cózar et al. 2005 75, 76
Squamous cell, esophagus Hsia et al. 2005 77
Squamous cell, vulva Sznurkowski et al. 2013 78
Open in a new tab

III. ACTIVATED NK CELLS LOCALIZE EFFICIENTLY AT TUMOR SITES

The densities of NK cells found in well-established tumors in most animal tumor models are also very low. NK-cell density can be increased somewhat by treatment with TLR agonists8,36,37 or with pro–NK-cell cytokines, particularly IL-2.38,39 This may be a result of improved extravasation or retention, survival, or proliferation of NK cells at the tumor sites (or of course, a combination of all). In contrast, very impressive NK-cell densities are seen in tumors after adoptive transfer of ex vivo IL-2–activated NK (A-NK) cells (Fig. 1).40–45 The density of NK cells reaches >500 cells mm−2 tumor tissue at 24 h after intravenous injection of 10×106 A-NK cells and increases to approximately 2,000 cells mm−2 by day 5.46 In this period of time, the density of A-NK cells in the tumor tissue is, on average, 20-fold higher than the density of A-NK cells in the surrounding normal lung tissue. Using the same assumption as above, this translates into E:T ratios from 1:4 to better than 1:1. The highest A-NK-cell densities are found in lung tumors, but significantly higher densities of A-NK cells in tumors compared to the surrounding normal tissues have been observed in liver, adrenal glands, spleen, bone marrow, brain, and ovary (Fig. 1).8 Interestingly, A-NK cells injected into the peritoneal cavity efficiently infiltrate tumors growing in the cavity; however, they seem to have some difficulty leaving the peritoneal cavity because lung tumors from animals receiving A-NK cells by the intraperitoneal (i.p.) route contain very few of the adoptively transferred cells at any time.47

FIG. 1. Accumulation of IL-2–activated NK (A-NK) cells selectively at tumor sites.

FIG. 1

Open in a new tab

Flow-sorted NKp46+ splenocytes from congenic Thy1.1+ C57BL/6 mice were cultured with IL-2 for 5 days and injected i.v. into C57BL/6 mice (Thy1.2+) with 9-day-old B16 tumors. Each mouse received 5 million A-NK cells. 30,000 IU Peg–IL-2 was injected i.p. every 12 h (max. six injections). Organs were removed at 72 h after injection of the A-NK cells and fresh frozen. Eight micron cryosections were all stained with PE-conjugated anti-Thy1.1 antibodies (NK cells begin to express Thy1 within 24 h of IL-2 activation). Some sections were also stained with FITC-conjugated anti-laminin antibodies. (A) DIC picture of lung tissue with multiple black-pigmented B16 melanoma metastases. (B) Fluorescent photomicrograph of the same sections as in (A), showing a dense accumulation of PE-Thy1.1+ A-NK cells (red dots) selectively in the black-pigmented metastases. White arrow points to a single PE-Thy1.1+ A-NK cell. (C) and (D) same as (A) and (B), respectively, but at higher magnification and with staining of laminin (green fluorescence in (D)). Note the strong preference of the A-NK cells for the tumor tissue. (E) and (F) show a DIC and a fluorescent picture, respectively, of laminin-stained ovarian tissue (green in (F)) with a black-pigmented B16 metastasis infiltrated by PE-Thy1.1+ A-NK cells. Bars in A–B = 200 μm, Bars in C–F = 100 μm.

To what extent these high intratumoral densities of A-NK cells are generated by a constant influx of A-NK cells or by proliferation of a few A-NK cells reaching the tumors (or both) is not fully elucidated. It is clear that proliferation of the A-NK cells, either in the tumor tissue or other places, is of major importance, because less than 250 A-NK cells mm−2 tumor tissue is found at 3 days after injection of irradiated (4 Gy) A-NK cells (Fig. 2). Furthermore, at 3 days after injection of non-irradiated, CFSE-labeled A-NK cells, hardly any of the A-NK cells contained enough CFSE for identification by fluorescence microscopy, indicating that the A-NK cells indeed continued to proliferate in vivo. The importance of cytokine-stimulation to maintain NK-cell proliferation is discussed later in this article.

FIG. 2. In vivo proliferation increases the number of transferred A-NK cells found at tumor sites.

FIG. 2

Open in a new tab

Flow-sorted NKp46+ splenocytes from congenic Thy1.1+, CD45.2+ and congenic Thy1.2+, CD45.1+ C57BL/6 mice were cultured with IL-2 for 5 days and injected i.v. into C57BL/6 mice (Thy1.2+,CD45.2) with 9-day-old B16 tumors. Before injection either the Thy1.1+ or the CD45.1+ A-NK cells were irradiated (450 rad). Each mouse received a mixture of 2.5 million Thy1.1+ and 2.5 million CD45.1+ A-NK cells. 30,000 IU Peg–IL-2 was injected i.p. every 12 h (max six injections). Organs were removed at 72 h after injection of the A-NK cells and were fresh frozen. (A) B16 lung metastasis from an animal receiving non-treated Thy1.1+ and irradiated CD45.1+ A-NK cells. While many Thy1.1+ A-NK cells (stained with FITC-anti-Thy1.1 antibodies) are infiltrating the tumor, only few of the irradiated CD45.1+ A-NK cells (stained with PE-anti-CD45.1 antibodies) can be observed. (B) Close-up of a B16 lung metastasis from an animal injected with irradiated Thy1.1+ and non-treated CD45.1+ A-NK cells. While few of the irradiated Thy1.1+ A-NK cells (stained with FITC-anti-Thy1.1 antibodies) are infiltrating the tumor, many non-irradiated CD45.1+ A-NK cells (stained with PE-anti-CD45.1 antibodies) can be observed throughout the tumor nodule. Bars = 50 μm.

IV. TUMOR-INFILTRATING A-NK CELLS HAVE ANTI-TUMOR ACTIVITY

Experimental lung metastases in most murine models are quite heterogeneous with respect to a variety of factors. This includes permissiveness to A-NK-cell infiltration.45 Thus, when comparing the fate of well-infiltrated lung tumors to that of poorly infiltrated lung tumors following adoptive transfer of A-NK cells by the intravenous (i.v.) route, it became clear that significant size reductions occurred only among the well-infiltrated tumors (Fig. 3).47 Furthermore, when the A-NK cells were injected using the i.p. route (hindering the A-NK cells in reaching any tumors except those growing in the i.p. cavity), a significant reduction of tumors in the i.p. cavity, but not of tumors in the lungs or any other organ, was observed. Thus, the ability of the A-NK cells to localize at tumor sites is not only impressive, but it is also a prerequisite for anti-tumor effect.

FIG. 3. Tumor-infiltrating A-NK cells eliminate tumor cells.

FIG. 3

Open in a new tab

Five million Thy1.1+ A-NK cells were injected into Thy1.2+ C57BL/6 mice bearing 9-day-old B16 melanoma lung metastases. 30,000 IU Peg–IL-2 was injected i.p. every 12 h (max. six injections). Organs were removed at 16 and 120 h after injection of the A-NK cells and were fresh frozen. (A) DIC picture of a B16 lung tumor from an animal injected with A-NK cells 16 h earlier. (B) Fluorescence picture of the tumor shown in (A) after staining with a 1:1 mixture of polyclonal rabbit anti-Tyrp114,15 and Pmel1716,17 antibodies (kindly provided by Dr. V. Hearing, NIH) and subsequently with FITC-conjugated anti-rabbit antibody. To reveal A-NK cells, the section was also stained with PE–anti-Thy1.1 antibodies. At this point in time, relatively few A-NK cells are found in the tumor, which is composed of B16 tumor cells (shoulder-to-shoulder). (C) DIC picture of B16 lung tumors from an animal injected with A-NK cells 120 h earlier. (D) Same area as in (C). Note that, while the tumor on the left is heavily infiltrated by A-NK cells (PE-Thy1.1+), which have almost replaced the tumor cells (FITCE+), the tumor on the right contains only few infiltrating A-NK cells but many tumor cells. Bars = 100 μm.

V. IN VIVO FUNCTION OF A-NK CELLS IS HIGHLY DEPENDENT ON CYTOKINE SUPPORT

Once activated with IL-2 or IL-15, in vitro–cultured A-NK cells become dependent on stimulation by these cytokines with respect to function, proliferation, and survival. Stimulation by just one of these cytokines is sufficient, regardless of which of them initially activated the NK cell. However, within just a few hours of deprivation of these cytokines, the proliferation of the A-NK cells slows down,46 and within less than 24 h, most of the A-NK cells have or will begin to undergo apoptosis. In vivo, lack of cytokine-support is evident by poor tumor-localization, loss of anti-tumor function and rapid disappearance of the A-NK cells from the recipient. Thus, in all of the studies mentioned above, substantial amounts of exogenous IL-2 were given to maintain the functionality of the A-NK cells, both with respect to tumor homing and anti-tumor effect. Due to the short plasma half-life of IL-2 (5–10 minutes), it is difficult, especially in animal models, to maintain the necessary high plasma levels of IL-2 by bolus injections of this cytokine. This problem can be solved by the use of pegylated IL-2 (Peg–IL-248–51) which, due to its greatly improved half-life (4–6 h48), needs to be injected just twice daily (approximately 30,000 IU bid,47). Unfortunately, the toxic side effects of Peg–IL-2 are enhanced in parallel with its beneficial effects on A-NK-cell homing and anti-tumor effect and the substantial toxicity of high-dose IL-2 treatment, especially the vascular leak syndrome, are well known.52–55 These side effects have greatly reduced enthusiasm for clinical usage of Peg–IL-2. The toxicity of Peg–IL-2 has, also in animal models, interfered with measurement of the anti-tumor effect of adoptive A-NK cell treatment in terms of improved survival, because 3 days of Peg-IL-2 treatment can be fatal for animals with high lung-tumor burdens. To circumvent this problem, NK cells have been modulated ex vivo, with vectors carrying genes for cytokines, which could enable the NK cells to produce their own IL-256 or to produce cytokines capable of synergizing with IL-2 in supporting the A-NK cells. One such cytokine is IL-12. A-NK cells pretreated with IL-12 or A-NK cells capable of IL-12–autostimulation via transgene IL-12 production need only 1/10–1/100 of the amount of IL-2 needed to maintain the same viability and functionality as A-NK cells that have not been stimulated with IL-12.57 This is likely due to the IL-12–induced expression of the IL-2 receptor alpha chain by the A-NK cells, enabling them to express the complete IL-2α-β-γ, high-affinity IL-2 receptor. Thus, by adoptive transfer of IL-12 transduced A-NK cells supported by just two injections of Peg–IL-2 (each of 3 × 104 IU) given on the same day as the A-NK cells, a significant prolongation of survival was observed in both 3-day and well-established 7-day models of B16 and MCA205 lung metastases.57,58 Although the A-NK-cell–produced IL-12 undoubtedly supported anti-tumor responses in addition to those generated by non–IL-12–producing A-NK cells, tumor homing by A-NK cells remained a very important and critical factor for the anti-tumor effect achieved by the IL-12 gene-transduced A-NK cells.58

VI. SURVIVAL OF A-NK CELLS IN VIVO

It is clear that both tumor homing and in vivo anti-tumor activity of A-NK cells are dependent on the continuous availability of IL-2 or IL-15, but it is less clear exactly which function(s) these cytokines support and which is most important. Possibly, they are causing changes not only in the NK cells but also in the tumor environment that are critical for the ability of the A-NK cells to sense the presence of the tumor cells, to extravasate, and to lyse the malignant cells. The answer may, however, be related to a more fundamental function, namely survival of the A-NK cells. It has long been known that lymph node–produced IL-15 is important for homeostasis of NK cells, i.e., if the NK cells are not frequently stimulated by IL-15, they rapidly die.59 Although a variety of cell types can produce IL-15 and present it in trans (which may be the most effective way of presenting IL-15 to NK cells60,61), it is likely that the amounts of IL-2 or IL-15 necessary to keep NK and A-NK cells alive are never being produced in tumors, since these are characterized by chronic inflammation (i.e., DAMPs rather than PAMPs) and expression NK cell-suppressive cytokines. Thus, within hours of arriving at a tumor site, the NK cells must leave again to find a source of IL-2 or IL-15 (e.g., the lymphoid tissues) or, maybe more likely, they rapidly die at the tumor site, many of them before they have had a chance to kill more than a few (if any) tumor cells. This hypothesis is supported by experiments showing that A-NK cells, transferred into tumor-bearing animals without any support by exogenous IL-2, are found at much higher densities in tumors gene-transduced to produce small amounts of IL-2 than in mock-transduced tumors.62 Likewise, adoptively transferred A-NK cells gene-transduced to produce just enough IL-2 to support their own survival in an intracrine fashion, i.e., with no detectable secretion of IL-2, were found in much higher numbers in tumors than mock-transduced A-NK cells (Fig. 4). Thus, it appears that, as long as the survival of the A-NK cells is ensured, they are able to traffic to and persist at tumor sites.

FIG. 4. Tumor-infiltrating A-NK cells depend on IL-2 for survival.

FIG. 4

Open in a new tab

GFP+ and GFP+,IL-2+ A-NK cells were produced by stable transduction of Mau-1 cells (a long-term A-NK-cell line (C57BL/6) developed in our lab) with adeno-associated virus containing the gene for GFP or the genes for both GFP and IL-2. (A) DIC picture of a B16 lung tumor from an animal injected i.v. with 5 million IL-2- and GFP-producing Mau-1 cells 72 h earlier. (B) Fluorescence picture of the tumor shown in (A) showing numerous GFP+ cells infiltrating the tumor. (C) DIC picture of a B16 lung tumor from an animal injected i.v. with 5 million GFP-producing Mau-1 cells 72 h earlier. (D) Fluorescence picture of the tumor shown in (A) showing very few GFP+ cells infiltrating the tumor. No exogenous IL-2 was given to support the injected Mau-1 A-NK cells. Bars = 100 μm.

VII. NK-CELL HOMING TO SITES OF INFECTION

Under steady conditions, the survival of NK cells seems to be maintained by lymphoid tissue–produced IL-15.59 However, as NK cells are believed to function as a first line of defense against especially intracellular infections, it would seem logical that, if cytokines are available at the site of infection, they are able to not only attract NK cells but also activate them and keep them alive. To test this, we established brain tumors in mice by injecting Her2-expressing D2F2/E2 mammary carcinoma cells into the cisterna magna (CM). Nine days later, when multiple tumors had formed in the brain parenchyma and the leptomeninges, targeted recombinant VSV with tropism only for Her2-expressing cells,63–66 were injected into the CM. Three days later, the brains were removed, and sections of brain tissue with tumor were stained with NKp46 antibody to reveal host NK cells. As shown in Fig. 5, an unprecedented high number of NK cells were found in the tumor tissue from animals receiving the targeted recombinant VSV compared to controls (which were not given the VSV). Furthermore, infiltration by the host NK cells was strictly confined to tumor tissue. Studies are ongoing to determine the extent to which the tumor-infiltrating NK cells are activated compared to NK cells found in the periphery and whether the high density of NK cells in the infected tumor tissue is caused by better survival (and possibly better proliferation as well) of NK cells arriving at this site or whether it is caused by an infection-induced influx of NK cells fast enough to outpace the loss due to intratumoral death of the NK cells. These studies show that, compared to the microenvironment of tumors, the milieu of infectious foci appear to provide the right conditions for the generation of high-density NK cell infiltrates.

FIG. 5. Viral infection induces a vigorous tumor infiltration by host NK cells.

FIG. 5

Open in a new tab

Mammary brain and leptomeningeal metastases were induced by injection of Her2neu expressing D2F2/E2 cells into the cisterna magna (CM) of Balb/c mice. Nine days later, Her2-targeting VSV were injected into the CM. Three days later, brains were removed and sections of brain were stained with NKp46 antibody to reveal host NK cells. (A) Fluorescence picture of a Hoechst 33342-stained brain section with a large D2F2/E2 tumor (the nuclei-dense lower half of the picture) from an animal receiving Her2-targeting VSV 3 days earlier. (B) Same area as (A), showing a high number of PE-NKp46+ host NK cells infiltrating the D2F2/E2 tumor (below the white line). Note the low density of host NK cells in the normal brain tissue (above the white line). (C) and (D) show an D2F2/E2 tumor from another animal receiving Her2 targeted VSV 3 days earlier (to the right of the white line). Note the dense cluster of host NK cells to the right and the absence of host NK cells from the adjacent normal brain tissue (to the left of the white out-line). (E) and (F) show an D2F2/E2 tumor (below white out-line) from a control animal, which did not receive the Her2-targeted virus. Note the very low number of host NK cells in the tumor tissue. Bars = 200 μm.

VIII. CONCLUSIONS AND PERSPECTIVES

It seems clear that the density of NK cells in the tumor microenvironment is a useful prognostic indicator in a variety of cancers, although we do not fully understand why this is the case. It is also clear that number of NK cells needed at tumor sites to allow the NK cells to influence tumor growth and viability are not spontaneously generated; however, such high densities of NK cells may be created if the “flavor” of the tumor microenvironment is changed to mimic that of a virally infected tissue. Alternatively, non-myeloablatic lymphodepletion has been shown to increase effector-cell survival and the anti-tumor effect in the setting of adoptive T-cell therapy of cancer in humans. It has been suggested that the rapid induction and secretion of cytokines in the host, in particular IL-7, to restore the lymphocytic homeostasis also supports the transferred T-effector cells. The same may be the case for NK cells. In fact, in a recent trial of adoptive NK-cell therapy given after non-myeloablatic lymphodepletion, high numbers of viable, transferred NK cells were detected in the blood of the recipients for at least 7 days (and in some patients for many weeks). Despite this positive finding, no objective responses were observed in any of the eight patients receiving this treatment.67 Thus, while the cytokine-response induced by the non-myeloablatic lymphodepletion may be sufficient to prolong survival of adoptively transferred NK cells as they circulate in the blood vascular system, it may not be sufficient to support those NK cells that leave the blood vascular system to infiltrate the malignant tissues. Likewise, the amount of IL-2 that was given to the patients in this study to support the transferred NK cells (750,000 IU kg−1 every 8 h for at least 2 days) was at least 50-fold lower than the amount of IL-2 that, in most animal models, is needed to successfully support adoptively transferred A-NK and other lymphokine-activated killer (LAK) cells.68 On this background, it is likely that the failure of adoptive NK cell therapy in this, as well as all other similar studies, is caused by insufficient availability, at the tumor sites, of cytokines capable of maintaining the NK cells' viability and anti-tumor functions. The relatively low doses of IL-2 and/or IL-15, which can be safely administered systemically, are far from sufficient to support intratumoral NK cells sufficiently. Therefore, better methods to ensure that the NK cells can produce their own IL-2 or IL-15 or strategies to increase the amounts of pro-NK cytokines in the microenvironment of tumors without simultaneously increasing their presence systemically, must be developed. In fact, we believe that similar strategies are needed to keep effector cells of T-cell origin alive at tumor sites. The superior survival of transferred CAR T cells, which incorporates the signaling portion of the 4-1BB receptor (ensuring that target-engagement of the CAR leads to IL-2 production by the CAR T cells69) strongly supports this notion. Because we strongly believe that the ultimate success of cellular immunotherapy will require the presence of both NK cells and T cells to eradicate both MHC class-I–positive and MHC class-I/ag–negative tumor cells, a strategy to ensure the survival of both NK cells and T cells selectively at the tumor site should lead to a substantial improvement in the efficacy of cell-mediated immunotherapy of disseminated cancer.

ACKNOWLEDGMENT

This research was supported by DoD BCRP grant BC101672P1/W81XWH-11-1-0125 (PHB) and in part by BC101672/W81XWH-11-1-0124 (YG). This project used the UPCI Cell and Tissue Imaging Facility (CTIF), Cytometry Facility (CF), and Vector Core (VC), which are supported, in part, by NIH award P30CA047904. We thank Dr. Ira Bergman for letting us use his D2F2/E2 tumor model and for his insightful reviews of our data and this manuscript We are grateful to Mrs. Lisa Bailey and Mrs. Jessica Poli for excellent technical and administrative help, respectively.

ABBREVIATIONS

A-NK

activated natural killer cells

Bid

twice per day

CAR

chimeric antigen receptor

CFSE

carboxyfluorescein succinimidyl ester

CM

cisterna magna

DAMP

damage-associated molecular pattern

E:T

effector-to-target ratio

Gy

gray

i.p.

intraperitoneal

i.v.

intravenous

IL

interleukin

IU

international unit

MHC

major histocompatibility complex

PAMP

pathogen-associated molecular pattern

Peg-IL-2

Pegylated interleukin-2

Poly I:C

polyinosinic:polycytidylic acid

Th1

T-helper-1

TLR

Toll-like receptor

VSV

vesicular stomatitis virus

REFERENCES

  • 1.Herberman RB, Nunn ME, Lavrin DH. Natural cytotoxic reactivity of mouse lymphoid cells against syngeneic acid allogeneic tumors. I. Distribution of reactivity and specificity. Int J Cancer. 1975 Aug 15;16(2):216–29. doi: 10.1002/ijc.2910160204. [DOI] [PubMed] [Google Scholar]
  • 2.Kiessling R, Klein E, Wigzell H. “Natural” killer cells in the mouse. I. Cytotoxic cells with specificity for mouse Moloney leukemia cells. Specificity and distribution according to genotype. Eur J I. 1975 Feb;5(2):112–7. doi: 10.1002/eji.1830050208. [DOI] [PubMed] [Google Scholar]
  • 3.Brunda MJ, Rosenbaum D, Stern L. Inhibition of experimentally-induced murine metastases by recombinant alpha interferon: correlation between the modulatory effect of interferon treatment on natural killer cell activity and inhibition of metastases. Int J Cancer. 1984 Sep 15;34(3):421–6. doi: 10.1002/ijc.2910340321. [DOI] [PubMed] [Google Scholar]
  • 4.Gorelik E, Herberman RB. Susceptibility of various strains of mice to urethan-induced lung tumors and depressed natural killer cell activity. J Nat Cancer Inst. 1981 Dec;67(6):1317–22. [PubMed] [Google Scholar]
  • 5.Gorelik E, Kedar E, Sredni B, Herberman R. In vivo anti-tumor effects of local adoptive transfer of mouse and human cultured lymphoid cells. Int J Cancer. 1981 Aug 15;28(2):157–64. doi: 10.1002/ijc.2910280208. [DOI] [PubMed] [Google Scholar]
  • 6.Hinuma S, Naruo K, Ootsu K, Houkan T, Shiho O, Tsukamoto K. Suppression of pulmonary tumour metastasis in mice by recombinant human interleukin-2: role of asialo GM1-positive cells. Immunology. 1987 Feb;60(2):173–9. [PMC free article] [PubMed] [Google Scholar]
  • 7.Basse P, Hokland P, Heron I, Hokland M. Fate of tumor cells injected into left ventricle of heart in BALB/c mice: role of natural killer cells. J Nat Cancer Inst. 1988 Jul 6;80(9):657–65. doi: 10.1093/jnci/80.9.657. [DOI] [PubMed] [Google Scholar]
  • 8.Yang Q, Goding SR, Hokland ME, Basse PH. Antitumor activity of NK cells. Immunol Res. 2006;36(1–3):13–25. doi: 10.1385/IR:36:1:13. [DOI] [PubMed] [Google Scholar]
  • 9.Vaquer S, Jorda J, Lopez de la Osa E, Alvarez de los Heros J, Lopez-Garcia N, Alvarez de Mon M. Clinical implications of natural killer (NK) cytotoxicity in patients with squamous cell carcinoma of the uterine cervix. Gynecol Oncol. 1990 Jan;36(1):90–2. doi: 10.1016/0090-8258(90)90114-z. [DOI] [PubMed] [Google Scholar]
  • 10.Dennert G, Landon C, Lord EM, Bahler DW, Frelinger JG. Lysis of a lung carcinoma by poly I:C-induced natural killer cells is independent of the expression of class I histocompatibility antigens. J Immunol. 1988 Apr 1;140(7):2472–5. [PubMed] [Google Scholar]
  • 11.Edwards BS, Borden EC, Smith-Zaremba K. Divergence in activation by poly I:C of human natural killer and killer cells. Cancer Immunol Immunother. 1982;13(3):158–63. doi: 10.1007/BF00205381. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Fresa KL, Korngold R, Murasko DM. Induction of natural killer cell activity of thoracic duct lymphocytes by polyinosinic-polycytidylic acid (poly(I:C)) or interferon. Cell Immunol. 1985 Apr 1;91(2):336–43. doi: 10.1016/0008-8749(85)90231-x. [DOI] [PubMed] [Google Scholar]
  • 13.Kiuchi Y. Strain difference in mouse natural killer activity augmented by mouse interferon and poly I.C. Jikken dobutsu Exper Anim. 1986 Jul;35(3):239–44. doi: 10.1538/expanim1978.35.3_239. [DOI] [PubMed] [Google Scholar]
  • 14.Murata A. The effect of activated pulmonary natural killer (NK) cells by poly I:C on the established pulmonary metastasis. Nihon Ika Daigaku Zasshi. 1989 Aug;56(4):383–91. doi: 10.1272/jnms1923.56.383. [DOI] [PubMed] [Google Scholar]
  • 15.Tokunaga T, Yamamoto S, Namba K. A synthetic single-stranded DNA, poly(dG,dC), induces interferon-alpha/beta and -gamma, augments natural killer activity, and suppresses tumor growth. Jap J Cancer Res. 1988 Jun;79(6):682–6. doi: 10.1111/j.1349-7006.1988.tb02222.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Won SJ, Lin MT. Alpha-methyl-p-tyrosine reduces poly I:C-induced augmenting interferon production and splenic natural killer cell activity in mice. Pharmacology. 1993 Jun;46(6):332–40. doi: 10.1159/000139063. [DOI] [PubMed] [Google Scholar]
  • 17.Markovic SN, Murasko DM. Anesthesia inhibits poly I:C induced stimulation of natural kill-er cell cytotoxicity in mice. Clin Immunol Immunopathol. 1990 Aug;56(2):202–9. doi: 10.1016/0090-1229(90)90141-c. [DOI] [PubMed] [Google Scholar]
  • 18.Beilin B, Martin FC, Shavit Y, Gale RP, Liebeskind JC. Suppression of natural killer cell activity by high-dose narcotic anesthesia in rats. Brain, Behav Immun. 1989 Jun;3(2):129–37. doi: 10.1016/0889-1591(89)90013-5. [DOI] [PubMed] [Google Scholar]
  • 19.Griffith CD, Rees RC, Platts A, Jermy A, Peel J, Rogers K. The nature of enhanced natural killer lymphocyte cytotoxicity during anesthesia and surgery in patients with benign disease and cancer. Ann Surg. 1984 Dec;200(6):753–8. doi: 10.1097/00000658-198412000-00014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Hsueh CM, Lorden JF, Hiramoto RN, Ghanta VK. Acquisition of enhanced natural killer cell activity under anesthesia. Life Sciences. 1992;50(26):2067–74. doi: 10.1016/0024-3205(92)90573-8. [DOI] [PubMed] [Google Scholar]
  • 21.Katzav S, Shapiro J, Segal S, Feldman M. General anesthesia during excision of a mouse tumor accelerates postsurgical growth of metastases by suppression of natural killer cell activity. Israel J Med Sci. 1986 May;22(5):339–45. [PubMed] [Google Scholar]
  • 22.Markovic SN, Murasko DM. Inhibition of induction of natural killer activity in mice by general anesthesia (Avertin): role of interferon. Clin Immunol Immunopathol. 1991 Aug;60(2):181–9. doi: 10.1016/0090-1229(91)90062-f. [DOI] [PubMed] [Google Scholar]
  • 23.Schantz SP, Romsdahl MM, Babcock GF, Nishioka K, Goepfert H. The effect of surgery on natural killer cell activity in head and neck cancer patients: in vitro reversal of a postoperatively suppressed immunosurveillance system. Laryngoscope. 1985 May;95(5):588–94. doi: 10.1288/00005537-198505000-00013. [DOI] [PubMed] [Google Scholar]
  • 24.Tartter PI, Martinelli G, Steinberg B, Barron D. Changes in peripheral T-cell subsets and natural-killer cytotoxicity in relation to colorectal cancer surgery. Cancer Detect Prevent. 1986;9(3–4):359–64. [PubMed] [Google Scholar]
  • 25.Tonnesen E, Mickley H, Grunnet N. Natural killer cell activity during premedication, anaesthesia and surgery. Acta Anaesthesiol Scand. 1983 Jun;27(3):238–41. doi: 10.1111/j.1399-6576.1983.tb01943.x. [DOI] [PubMed] [Google Scholar]
  • 26.Tai LH, de Souza CT, Belanger S, Ly L, Alkayyal AA, Zhang J, Rintoul JL, Ananth AA, Lam T, Breitcach CJ, Falls TJ, Kirn DH, Bell JC, Makrigiannis AP, Auer RA. Preventing postoperative metastatic disease by inhibiting surgery-induced dysfunction in natural killer cells. Cancer Res. 2013 Jan 1;73(1):97–107. doi: 10.1158/0008-5472.CAN-12-1993. [DOI] [PubMed] [Google Scholar]
  • 27.Tai LH, Zhang J, Scott KJ, de Souza CT, Alkayyal AA, Ananth AA, Sahi S, Adair RA, Mahmoud AB, Sad S, Bell JC, Makrigiannis AP, Melcher AA, Auer RC. Perioperative influenza vaccination reduces postoperative metastatic disease by reversing surgery-induced dysfunction in natural killer cells. Clin Cancer Res. 2013 Sep 15;19(18):5104–15. doi: 10.1158/1078-0432.CCR-13-0246. [DOI] [PubMed] [Google Scholar]
  • 28.Kondo E, Koda K, Takiguchi N, Oda K, Seike K, Ishizuka M, Miyazaki M. Preoperative natural killer cell activity as a prognostic factor for distant metastasis following surgery for colon cancer. Digestive Surgery. 2003;20(5):445–51. doi: 10.1159/000072714. [DOI] [PubMed] [Google Scholar]
  • 29.Barlozzari T, Reynolds CW, Herberman RB. In vivo role of natural killer cells: involvement of large granular lymphocytes in the clearance of tumor cells in anti-asialo GM1-treated rats. J Immunol. 1983 Aug;131(2):1024–7. [PubMed] [Google Scholar]
  • 30.Hanna N. Role of natural killer cells in control of cancer metastasis. Cancer Metastasis Revs. 1982;1(1):45–64. doi: 10.1007/BF00049480. [DOI] [PubMed] [Google Scholar]
  • 31.Introna M, Mantovani A. Natural killer cells in human solid tumors. Cancer Metastasis Revs. 1983;2(4):337–50. doi: 10.1007/BF00048566. [DOI] [PubMed] [Google Scholar]
  • 32.Guerra N, Tan YX, Joncker NT, Choy A, Gallardo F, Xiong N, Knoblaugh S, Cado D, Greenberg NM, Raulet DH. NKG2D-deficient mice are defective in tumor surveillance in models of spontaneous malignancy. Immunity. 2008 Apr;28(4):571–80. doi: 10.1016/j.immuni.2008.02.016. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Coca S, Perez-Piqueras J, Martinez D, Colmenarejo A, Saez MA, Vallejo C, Martos JA, Moreno M. The prognostic significance of intratumoral natural killer cells in patients with colorectal carcinoma. Cancer. 1997 Jun 15;79(12):2320–8. doi: 10.1002/(sici)1097-0142(19970615)79:12<2320::aid-cncr5>3.0.co;2-p. [DOI] [PubMed] [Google Scholar]
  • 34.Villegas FR, Coca S, Villarrubia VG, Jiménez R, Chillón MJ, Jareño J, Zuil M, Callol L. Prognostic significance of tumor infiltrating natural killer cells subset CD57 in patients with squamous cell lung cancer. Lung Cancer. 2002 Jan;35(1):23–8. doi: 10.1016/s0169-5002(01)00292-6. [DOI] [PubMed] [Google Scholar]
  • 35.Vaquero J, Zurita M, Coca S. Clinical behavior in metastatic brain disease is not influenced by the immunological defense mediated by CD57 NK-Cells. Int J Surg Oncol. 2012;2012:834852. doi: 10.1155/2012/834852. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Smyth MJ, Kelly JM, Baxter AG, Korner H, Sedgwick JD. An essential role for tumor necrosis factor in natural killer cell-mediated tumor rejection in the peritoneum. J Exper Med. 1998 Nov 2;188(9):1611–9. doi: 10.1084/jem.188.9.1611. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Chew V, Tow C, Huang C, Bard-Chapeau E, Copeland NG, Jenkins NA, Weber A, Lim KH, Toh HC, Heikenwalder M, Ng IO, Nardin A, Abastado JP. Toll-like receptor 3 expressing tumor parenchyma and infiltrating natural killer cells in hepatocellular carcinoma patients. J Nat Cancer Inst. 2012 Dec 5;104(23):1796–807. doi: 10.1093/jnci/djs436. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Hokland M, Kjaergaard J, Kuppen PJ, Nannmark U, Agger R, Hokland P, Basse P. Endogenous and adoptively transferred A-NK and T-LAK cells continuously accumulate within murine metastases up to 48 h after inoculation. In Vivo. 1999 May-Jun;13(3):199–204. [PubMed] [Google Scholar]
  • 39.Wendel M, Galani IE, Suri-Payer E, Cerwenka A. Natural killer cell accumulation in tumors is dependent on IFN-gamma and CXCR3 ligands. Cancer Res. 2008 Oct 15;68(20):8437–45. doi: 10.1158/0008-5472.CAN-08-1440. [DOI] [PubMed] [Google Scholar]
  • 40.Basse P, Goldfarb RH. Localization of immune effector cells to tumor metastases. Immunol Ser. 1994;61:149–58. [PubMed] [Google Scholar]
  • 41.Basse P, Herberman RB, Nannmark U, Johansson BR, Hokland M, Wasserman K, Goldfarb RH. Accumulation of adoptively transferred adherent, lymphokine-activated killer cells in murine metastases. J Exper Med. 1991 Aug 1;174(2):479–88. doi: 10.1084/jem.174.2.479. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Basse PH. Tissue distribution and tumor localization of effector cells in adoptive immunotherapy of cancer. APMIS Suppl. 1995;55:1–28. [PubMed] [Google Scholar]
  • 43.Basse PH, Nannmark U, Johansson BR, Herberman RB, Goldfarb RH. Establishment of cell-to-cell contact by adoptively transferred adherent lymphokine-activated killer cells with metastatic murine melanoma cells. J Nat Cancer Inst. 1991 Jul 3;83(13):944–50. doi: 10.1093/jnci/83.13.944. [DOI] [PubMed] [Google Scholar]
  • 44.Vujanovic NL, Basse P, Herberman RB, Whiteside TL. Antitumor functions of natural killer cells and control of metastases. Methods. 1996 Apr;9(2):394–408. doi: 10.1006/meth.1996.0044. [DOI] [PubMed] [Google Scholar]
  • 45.Yang Q, Goding S, Hagenaars M, Carlos T, Albertsson P, Kuppen P, Nannmark U, Hokland ME, Basse PH. Morphological appearance, content of extracellular matrix and vascular density of lung metastases predicts permissiveness to infiltration by adoptively transferred natural killer and T cells. Cancer Immunol Immunother. 2006 Jun;55(6):699–707. doi: 10.1007/s00262-005-0043-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Basse PH, Goldfarb RH, Herberman RB, Hokland ME. Accumulation of adoptively transferred A-NK cells in murine metastases: kinetics and role of interleukin-2. In Vivo. 1994 Jan-Feb;8(1):17–24. [PubMed] [Google Scholar]
  • 47.Yang Q, Hokland ME, Bryant JL, Zhang Y, Nannmark U, Watkins SC, Goldfarb RH, Herberman RB, Basse PH. Tumor-localization by adoptively transferred, interleukin-2-activated NK cells leads to destruction of well-established lung metastases. Int J Cancer. 2003 Jul 1;105(4):512–9. doi: 10.1002/ijc.11119. [DOI] [PubMed] [Google Scholar]
  • 48.Katre NV, Knauf MJ, Laird WJ. Chemical modification of recombinant interleukin 2 by polyethylene glycol increases its potency in the murine Meth A sarcoma model. Proc Nat Acad Sci U S A. 1987 Mar;84(6):1487–91. doi: 10.1073/pnas.84.6.1487. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.Zimmerman RJ, Aukerman SL, Katre NV, Winkelhake JL, Young JD. Schedule dependency of the antitumor activity and toxicity of polyethylene glycol-modified interleukin 2 in murine tumor models. Cancer Res. 1989 Dec 1;49(23):6521–8. [PubMed] [Google Scholar]
  • 50.Meyers FJ, Paradise C, Scudder SA, Goodman G, Konrad M. A phase I study including pharmacokinetics of polyethylene glycol conjugated interleukin-2. Clin Pharmacol Ther. 1991 Mar;49(3):307–13. doi: 10.1038/clpt.1991.33. [DOI] [PubMed] [Google Scholar]
  • 51.Yang JC, Schwarz SL, Perry-Lalley DM, Rosenberg SA. Murine studies using polyethylene glycol-modified recombinant human interleukin 2 (PEG-IL-2): antitumor effects of PEG-IL2 alone and in combination with adoptive cellular transfer. Lymphokine Cytokine Res. 1991 Dec;10(6):475–80. [PubMed] [Google Scholar]
  • 52.Cotran RS, Pober JS, Gimbrone MA, Jr, Springer TA, Wiebke EA, Gaspari AA, Rosenberg SA, Lotze MT. Endothelial activation during interleukin 2 immunotherapy. A possible mechan-ism for the vascular leak syndrome. J Immunol. 1988 Mar 15;140(6):1883–8. [PubMed] [Google Scholar]
  • 53.Pockaj BA, Yang JC, Lotze MT, Lange JR, Spencer WF, Steinberg SM, Topalian SL, Schwartzentruber DJ, White DE, Rosenberg SA. A prospective randomized trial evaluating colloid versus crystalloid resuscitation in the treatment of the vascular leak syndrome associated with interleukin-2 therapy. J Soc Biol Ther. 1994 Jan;15(1):22–8. doi: 10.1097/00002371-199401000-00003. [DOI] [PubMed] [Google Scholar]
  • 54.Baluna R, Rizo J, Gordon BE, Ghetie V, Vitetta ES. Evidence for a structural motif in toxins and interleukin-2 that may be responsible for binding to endothelial cells and initiating vascular leak syndrome. Proc Nat Acad Sci U S A. 1999 Mar 30;96(7):3957–62. doi: 10.1073/pnas.96.7.3957. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 55.Samlowski WE, Kondapaneni M, Tharkar S, McGregor JR, Laubach VE, Salvemini D. Endothelial nitric oxide synthase is a key mediator of interleukin-2-induced hypotension and vascular leak syndrome. J Immunother. 2011 Jun;34(5):419–27. doi: 10.1097/CJI.0b013e31821dcb50. [DOI] [PubMed] [Google Scholar]
  • 56.Nagashima S, Mailliard R, Kashii Y, Reichert TE, Herberman RB, Robbins P, Whiteside TL. Stable transduction of the interleukin-2 gene into human natural killer cell lines and their phenotypic and functional characterization in vitro and in vivo. Blood. 1998 May 15;91(10):3850–61. [PubMed] [Google Scholar]
  • 57.Goding S, Yang Q, Mi Z, Robbins PD, Basse PH. Targeting of products of genes to tumor sites using adoptively transferred A-NK and T-LAK cells. Cancer Gene Therapy. 2007 May;14(5):441–50. doi: 10.1038/sj.cgt.7701019. [DOI] [PubMed] [Google Scholar]
  • 58.Goding SR, Yang Q, Knudsen KB, Potter DM, Basse PH. Cytokine gene therapy using adenovirally transduced, tumor-seeking activated natural killer cells. Human Gene Therapy. 2007 Aug;18(8):701–11. doi: 10.1089/hum.2007.052. [DOI] [PubMed] [Google Scholar]
  • 59.Cooper MA, Bush JE, Fehniger TA, VanDeusen JB, Waite RE, Liu Y, Aguila HL, Caligiuri MA. In vivo evidence for a dependence on interleukin 15 for survival of natural killer cells. Blood. 2002 Nov 15;100(10):3633–8. doi: 10.1182/blood-2001-12-0293. [DOI] [PubMed] [Google Scholar]
  • 60.Burkett PR, Koka R, Chien M, Chai S, Boone DL, Ma A. Coordinate expression and trans presentation of interleukin (IL)-15Ralpha and IL-15 supports natural killer cell and memory CD8+ T cell homeostasis. J Exper Med. 2004 Oct 4;200(7):825–34. doi: 10.1084/jem.20041389. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 61.Vujanovic L, Szymkowski DE, Alber S, Watkins SC, Vujanovic NL, Butterfield LH. Virally infected and matured human dendritic cells activate natural killer cells via cooperative activity of plasma membrane-bound TNF and IL-15. Blood. 2010 Jul 29;116(4):575–83. doi: 10.1182/blood-2009-08-240325. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 62.Katakam AK, Yang Q, Goding S, Kumar P, Basse PH. The RheoSwitch[reg] therapeutic system precisely regulates IL-2 expression in melanoma cells and supports survival of NK cells at the tumor site. Mol Ther. 2006;13(S1):S103–S. 270. 05//print. [Google Scholar]
  • 63.Bergman I, Whitaker-Dowling P, Gao Y, Griffin JA. Preferential targeting of vesicular stomatitis virus to breast cancer cells. Virology. 2004 Dec 5;330(1):24–33. doi: 10.1016/j.virol.2004.06.048. [DOI] [PubMed] [Google Scholar]
  • 64.Gao Y, Whitaker-Dowling P, Watkins SC, Griffin JA, Bergman I. Rapid adaptation of a recombinant vesicular stomatitis virus to a targeted cell line. J Virol. 2006 Sep;80(17):8603–12. doi: 10.1128/JVI.00142-06. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 65.Bergman I, Griffin JA, Gao Y, Whitaker-Dowling P. Treatment of implanted mammary tumors with recombinant vesicular stomatitis virus targeted to Her2/neu. Int J Cancer. 2007 Jul 15;121(2):425–30. doi: 10.1002/ijc.22680. [DOI] [PubMed] [Google Scholar]
  • 66.Gao Y, Whitaker-Dowling P, Griffin JA, Barmada MA, Bergman I. Recombinant vesicular stomatitis virus targeted to Her2/neu combined with anti-CTLA4 antibody eliminates implanted mammary tumors. Cancer Gene Therapy. 2009 Jan;16(1):44–52. doi: 10.1038/cgt.2008.55. [DOI] [PubMed] [Google Scholar]
  • 67.Parkhurst MR, Riley JP, Dudley ME, Rosenberg SA. Adoptive transfer of autologous natural killer cells leads to high levels of circulating natural killer cells but does not mediate tumor regression. Clin Cancer Res. 2011 Oct 1;17(19):6287–97. doi: 10.1158/1078-0432.CCR-11-1347. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 68.Mule JJ, Yang J, Shu S, Rosenberg SA. The anti-tumor efficacy of lymphokine-activated killer cells and recombinant interleukin 2 in vivo: direct correlation between reduction of established metastases and cytolytic activity of lymphokine-activated killer cells. J Immunol. 1986 May 15;136(10):3899–909. [PubMed] [Google Scholar]
  • 69.Song DG, Ye Q, Carpenito C, Poussin M, Wang LP, Ji C, Figini M, June CH, Coukos G, Powell DJ., Jr. In vivo persistence, tumor localization, and antitumor activity of CAR-engineered T cells is enhanced by costimulatory signaling through CD137 (4-1BB) Cancer Res. 2011 Jul 1;71(13):4617–27. doi: 10.1158/0008-5472.CAN-11-0422. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 70.Taketomi A, Shimada M, Shirabe K, Kajiyama K, Gion T, Sugimachi K. Natural killer cell activity in patients with hepatocellular carcinoma: a new prognostic indicator after hepatectomy. Cancer. 1998 Jul 1;83(1):58–63. doi: 10.1002/(sici)1097-0142(19980701)83:1<58::aid-cncr8>3.0.co;2-a. [DOI] [PubMed] [Google Scholar]
  • 71.Ishigami S, Natsugoe S, Tokuda K, Nakajo A, Che X, Iwashige H, Aridome K, Hokita S, Aikou T. Prognostic value of intratumoral natural killer cells in gastric carcinoma. Cancer. 2000 Feb 1;88(3):577–83. [PubMed] [Google Scholar]
  • 72.Takanami I, Takeuchi K, Giga M. The prognostic value of natural killer cell infiltration in resected pulmonary adenocarcinoma. J Thoracic Cardiovascular Surg. 2001 Jun;121(6):1058–63. doi: 10.1067/mtc.2001.113026. [DOI] [PubMed] [Google Scholar]
  • 73.Takeuchi H, Maehara Y, Tokunaga E, Koga T, Kakeji Y, Sugimachi K. Prognostic significance of natural killer cell activity in patients with gastric carcinoma: a multivariate analysis. Am J Gastroenterol. 2001 Feb;96(2):574–8. doi: 10.1111/j.1572-0241.2001.03535.x. [DOI] [PubMed] [Google Scholar]
  • 74.Lowdell MW, Craston R, Samuel D, Wood ME, O'Neill E, Saha V, Prentice HG. Evidence that continued remission in patients treated for acute leukaemia is dependent upon autologous natural killer cells. Brit J Haematol. 2002 Jun;117(4):821–7. doi: 10.1046/j.1365-2141.2002.03495.x. [DOI] [PubMed] [Google Scholar]
  • 75.Cozar JM, Canton J, Tallada M, Concha A, Cabrera T, Garrido F, Ruiz-Cabello Osuna F. Analysis of NK cells and chemokine receptors in tumor infiltrating CD4 T lymphocytes in human renal carcinomas. Cancer Immunol Immunother. 2005 Sep;54(9):858–66. doi: 10.1007/s00262-004-0646-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 76.Donskov F. Interleukin-2 based immunotherapy in patients with metastatic renal cell carcinoma. Danish Med Bull. 2007 Nov;54(4):249–65. [PubMed] [Google Scholar]
  • 77.Hsia JY, Chen JT, Chen CY, Hsu CP, Miaw J, Huang YS, Yang CY. Prognostic significance of intratumoral natural killer cells in primary resected esophageal squamous cell carcinoma. Chang Gung Med J. 2005 May;28(5):335–40. [PubMed] [Google Scholar]
  • 78.Sznurkowski JJ, Zawrocki A, Biernat W. Subtypes of cytotoxic lymphocytes and natural killer cells infiltrating cancer nests correlate with prognosis in patients with vulvar squamous cell carcinoma. Cancer Immunol Immunother. 2014 Mar;63(3):297–303. doi: 10.1007/s00262-013-1511-x. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES