Search

Scientific discoveries that provide strong evidence of antitumor effects in preclinical models often encounter significant delays before being tested in patients with cancer. While some of these delays have a scientific basis, others do not. We need to do better. Innovative strategies need to move into early stage clinical trials as quickly as it is safe, and if successful, these therapies should efficiently obtain regulatory approval and widespread clinical application. In late 2009 and 2010 the Society for Immunotherapy of Cancer (SITC), convened an "Immunotherapy Summit" with representatives from immunotherapy organizations representing Europe, Japan, China and North America to discuss collaborations to improve development and delivery of cancer immunotherapy. One of the concepts raised by SITC and defined as critical by all parties was the need to identify hurdles that impede effective translation of cancer immunotherapy. With consensus on these hurdles, international working groups could be developed to make recommendations vetted by the participating organizations. These recommendations could then be considered by regulatory bodies, governmental and private funding agencies, pharmaceutical companies and academic institutions to facilitate changes necessary to accelerate clinical translation of novel immune-based cancer therapies. The critical hurdles identified by representatives of the collaborating organizations, now organized as the World Immunotherapy Council, are presented and discussed in this report. Some of the identified hurdles impede all investigators; others hinder investigators only in certain regions or institutions or are more relevant to specific types of immunotherapy or first-in-humans studies. Each of these hurdles can significantly delay clinical translation of promising advances in immunotherapy yet if overcome, have the potential to improve outcomes of patients with cancer. © 2011 Fox et al; licensee BioMed Central Ltd.

Cancer stem cells (CSC) are thought to be responsible for tumor initiation and tumor regeneration after chemotherapy. Previously, we showed that chemotherapy of non–small cell lung cancer (NSCLC) cells lines can select for outgrowth of highly tumorigenic and metastatic CSCs. The high malignancy of lung CSCs was associated with an efficient cytokine network. In this study, we provide evidence that blocking stem cell factor (SCF)–c-kit signaling is sufficient to inhibit CSC proliferation and survival promoted by chemotherapy. CSCs were isolated from NSCLC cell lines as tumor spheres under CSC-selective conditions and their stem properties were confirmed. In contrast to other tumor cells, CSCs expressed c-kit receptors and produced SCF. Proliferation of CSCs was inhibited by SCF-neutralizing antibodies or by imatinib (Gleevec), an inhibitor of c-kit. Although cisplatin treatment eliminated the majority of tumor cells, it did not eliminate CSCs, whereas imatinib or anti-SCF antibody destroyed CSCs. Significantly, combining cisplatin with imatinib or anti-SCF antibody prevented the growth of both tumor cell subpopulations. Our findings reveal an important role for the SCF–c-kit signaling axis in self-renewal and proliferation of lung CSCs, and they suggest that SCF–c-kit signaling blockade could improve the antitumor efficacy of chemotherapy of human NSCLC. Cancer Res; 70(1); 338–46

To study the cellular structures involved in NK and lymphokine-activated killer (LAK) cell function, we have produced a panel of mAbs that modulate the cytolytic function of a population of cells with LAK activity that derive from large granular lymphocyte (LGL)/NK cells (adherent LAK [A-LAK] cells). In this report, we describe an mAb (3.2.3; IgG1k) that recognizes a triggering structure that is expressed on rat LGL/NK cells and A-LAK cells. This epitope is also expressed on polymorphonuclear leukocytes (PMN). The expression of the epitope identified by mAb 3.2.3 increased progressively on A-LAK cells after culture in the presence of rIL-2. mAb 3.2.3 enhanced the cytolytic activity of NK and A-LAK cells against FcR+ target cells, but not FcR- target cells. However, this effect was not induced by F(ab')2 fragments of 3.2.3. This antibody also induced the release of N-alpha-benzyloxycarbonyl-L-lysine thiobenzy esteresterase by A-LAK cells. These data suggest that the epitope identified by mAb 3.2.3 is on a triggering structure expressed on rat NK cells and A-LAK cells. The expression of the epitope recognized by mAb 3.2.3 on LGL/NK cells and PMN suggests that this structure may be analogous to that identified by the anti-CD16 (-FcR) mAbs. However, the molecule immunoprecipitated by mAb 3.2.3 was a 60-kD dimer composed of two 30-kD chains. These data suggest that mAb 3.2.3 recognizes a unique triggering structure. As mAb 3.2.3 is the first antibody recognizing a determinant with functional significance, selectively expressed on both rat NK cells and A-LAK cells, it will be a useful tool for the study of NK cell ontogeny and function, and the development of cells with LAK activity from the NK cell compartment.

A simple method for the purification and rapid expansion of large granular lymphocytes into cells with efficient broad antitumor cytotoxicity after stimulation by human rIL-2 is described. Nylon-wool nonadherent splenic mononuclear leukocytes from Fischer 344 rats were cultured in medium containing 1,000 U/ml rIL-2. The initial response of a small subpopulation of cells (less than 2%) to rIL-2 was their adherence to the plastic surface. This response was noted as soon as 2 h after addition of rIL-2. 2-h rIL-2-activated plastic adherent lymphocytes were 90-98% LGL, expressed surface markers characteristic of rat NK cells (OX8 [CD8]+, asialo GM1, laminin+, OX19 [CD5]-, R1-3B3 [CD5]-, W3/25 [CD4]-, OX39 [CD25]-, Ia-, and Ig-), and expressed very high levels of cytotoxicity against YAC-1 target cells. In addition to the above markers, plastic-adherent LGLs obtained at 24, 48, or 72 h progressively expressed Ia surface antigens, but were not phagocytic and contained less than 1% monocytes/macrophages by morphology. When 24- or 48-h plastic-adherent LGL/NK cells were cultured over 3-4 d in rIL-2, the cells expanded between 30- and 100-fold, reaching densities between 2-3 X 10(6) cells/ml. These rapidly expanding LGL/NK cells also generated very high levels of LAK activity (including lysis of fresh NK-resistant solid tumor cells), expressed a phenotype characteristic of activated rat NK/LAK cells, and incorporated [3H]TdR into DNA. This technique not only provides a novel method for the purification of LGL/NK cells for in vitro studies but also provides a means for the rapid expansion of highly purified cells with high levels of broad antitumor (LAK) cytotoxicity.

In this report we have used highly purified populations of natural killer (NK) cells: large granular lymphocytes (LGL). This study demonstrates that freshly isolated and interleukin 2-cultured LGL produce the specific neutral serine protease, plasminogen activator (PA). We have found that the enzyme is expressed in both an extracellular form as well as in a cell-associated form. Upon subcellular distribution the latter form of the enzyme is associated with a cell-surface membrane-enriched fraction. LGL PA exists in multiple molecular weight forms ranging from 100,000 to 26,000. Interferon (IFN), the major positive regulator of NK cytolytic activity, caused a substantial enhancement of cell-associated, but not extracellular, PA. In contrast, LGL isolated from patients with Chediak-Higashi syndrome, who are known to be defective in NK activity, displayed low PA activity, altered morphology, and low NK killing relative to LGL isolated from normal donors. The possible role of LGL PA in the lysis of tumor cells by NK cells, either directly or indirectly, is discussed.

Augmentation of natural killer (NK) activity by influenza A/PC and HSV-1 viruses appears to be caused by the induction of interferon (IFN) within the NK cell population itself. These viruses induced high levels of IFN production by human large granular lymphocytes (LGL) that could be readily isolated from peripheral blood by Percoll density gradients. These LGL, which have been previously shown to account for and to be highly associated with endogenous NK activity, became augmented in their lytic function during the 18-h period that IFN was induced. Non-LGL helper cells did not appear to be required in the NK-IFN system (either T cells, B cells, or monocytes). Removal of these latter cell types by treatment with OKT3 plus complement, anti-IgM plus complement, or preincubation with silica or carrageenan had no effect on the ability of LGL to respond to the viruses. Production of IFN was also detected, albeit at lower levels, from monocytes cultured for 18 h with viruses, but no cytotoxic activity was induced. On the other hand, T cells, even in the presence of monocytes, showed neither property, and longer cultures, with virus up to 4 d, still did not alter the pattern. The IFN produced by both LGL and monocytes were predominantly IFN-alpha, as assessed by neutralization assays with antisera to IFN-alpha, -beta, and -gamma. In an individual with detectable serum antibodies to influenza A/PC, however, the IFN induced in LGL appeared to be gamma, presumably because of specific recognition of the virus. These data suggest an efficient positive self-regulatory mechanism in NK cells that may be readily switched on by viruses.

Spleen cells from W/Fu rats 40 days or more after immunization with a syngeneic Gross virus-induced leukemia were unreactive in direct cytotoxic assays. Incubation of these immune cells at 37 degrees C for 12 hr or longer, in the absence of antigen, resulted in the appearance of specific cytotoxic reactivity. Other lymphoid cells from the immune rats also were activated upon in vitro incubation, but to a lesser extent. Experiments were performed to define the necessary conditions and the mechanism for the in vitro incubation. Activation was temperature dependent, occurring at 37 degrees C but not at 4 degrees C. Immune serum suppressed the activation, but normal rat serum also had some inhibitory activity. Passage of immune cells through a nylon column, before preincubation, prevented activation. In contrast, exposure to nylon after preincubation did not remove cytotoxic reactivity. These findings demonstrate the reversal of a central inhibition of immune cell activity. The explanations offered for this phenomenon included change in surface characteristics of the immune cells during in vitro incubation, and the possible need for an adherent helper cell.

Studies were performed to determine the basis for low NK activity in the spleens of SJL/J mice. In contrast to substantial boosting by IL-2 or IFN of NK activity of spleen cells from high NK-reactive strains of mice, no detectable effect of these cytokines on SJL spleen cells was seen. When SJL spleen cells were cultured for 7 days in the presence of IL-2, the frequency of proliferating cells was comparable to that of other strains of mice. However, the SJL spleen cells showed a frequency of NK-cell progenitors which was at least 50 times lower than that of the CBA/J spleen cells. In addition to having no detectable effector-cell activity, the cultures of SJL spleen cells contained suppressor cells which were able to inhibit the NK activity of spleen cells from other strains of mice. These suppressor cells did not adhere to plastic or nylon wool and were found in low-density fractions after Percoll density gradient centrifugation. The cultured SJL spleen cells had a high percentage of cells capable of binding to NK-susceptible target cells, which was similar to that seen in lytic cultures from other strains of mice. Thus, the suppressor activity may be attributable to competitive inhibition of the interaction of effector cells with target cells. Although several of the characteristics of suppressor cells were similar to those of cultured effector cells, they may not represent inactive or pre-NK cells since their progenitors were Thy1+ and AsialoGM1- whereas the progenitors of cultured effector cells in high NK strains were Thy1- and GM1+. The ability to eliminate the progenitors of the suppressor cells by pretreatment of spleen cells with anti-Thy1 plus C' also suggested that the low or undetectable NK activity of the cultured SJL cells is not attributable to suppression but may be due to an inherent deficit of NK cells.

The metastatic ability of 3LL tumor following in vitro irradiation with ultraviolet (u.v.) light was studied. Tumor cells were exposed to two courses of u.v.-irradiation (3LL × 2u.v. cells) and after two weeks of culture they were inoculated intravenously (i.v.) into syngeneic mice. These cells produced significantly fewer pulmonary metastases than the untreated population. In addition, intrafootpad (i.f.p.) injections of 3LL × 2u.v. cells into immunocompetent animals induced tumors only in 40 per cent of recipients. Interestingly, in normal mice with progressively growing 3LL × 2u.v. tumors, the formation of spontaneous pulmonary metastases was prevented, whereas metastatic foci were observed in 70 per cent of the nude recipients. The metastatic properties of u.v.-treated tumor cells were further analysed by using individual clones with varying immunogenicity. We found that variants with augmented immunogenicity also showed a parallel decrease in metastatic potential. Studies on H-2 antigen expression in different clones revealed that immunogenic and low metastatic variants expressed levels of H-2 antigens higher than the tumorigenic and metastatic clones. Finally, by using cyclophosphamide (Cy) treatment and adoptive transfer of immune spleen cells we were able to eradicate macroscopic 3LL pulmonary metastasis. These results demonstrate that the decrease of metastatic ability in u.v.-treated cells was mainly due to an increase in their immunogenicity and H-2 antigen expression.

EVIDENCE has accumulated which stresses the importance of lymphokines in cell-mediated immunity both in vivo and in vitro1. But, the processes leading to release of lymphokines by antigen-activated lymphocytes, the mode of action of these lymphokines and their exact role in cell-mediated immunity are still not well defined. It has, however, been recognised that in vitro activation of immune lymphocytes by soluble antigens is macrophage dependent and that this activation is under the control of gene products of the major histocompatibility complex (H–2 in the mouse)2. Using a Moloney murine sarcoma virus (MSV)-induced tumour system, we have shown, in agreement with reports of some other systems3,4, that immune T lymphocytes required macrophages for migration inhibition factor (MIF) production and, in addition, only histocompatible macrophages could assist immune T lymphocytes for MIF release after stimulation with soluble tumour-associated antigens5. We report here that gene products of the H–2 complex regulated the macrophage-immune lymphocyte interaction for MIF release when soluble tumour extracts were used as the source of antigen. In contrast, when intact tumour cells were used as the source of antigen, a macrophage requirement was not detectable. Moreover, the direct lymphocyte-tumour cell interaction for MIF release was not under H–2 restriction, since allogenic as well as syngeneic tumour cells could activate MIF release from immune lymphocytes. These findings suggest the existence of at least two different pathways for MIF production, reflecting perhaps the activation of two distinct sub-populations of T lymphocytes or the activation of T lymphocytes at two different stages of differentiation.

Lethally irradiated mice pretreated with a wide range of Adriamycin (ADM) doses were tested for their in vivo antitumor response against transplanted lymphoma cells. Tumor growth, as assessed by 125IUdR uptake, was markedly impaired by treatment with ADM by a variety of different administration schedules. This ADM-induced antitumor response was largely dose-dependent, occurred regardless of route of administration, and was detectable as late as 15-30 days following drug treatment. No genetic restriction could be found to regulate the development of this response, since tumor growth inhibition occurred in syngeneic as well as allogeneic tumor-host combinations. ADM-induced antitumor response did not appear to be due to direct antitumor action by the drug, but rather to some interaction with host immune mechanisms. Antimacrophage agents, such as silica or carrageenan, abrogated the response. The possible implication of different cells as effectors of this response is discussed.

Publisher Summary This chapter focuses on mechanisms underlying the depression of natural killer (NK) activity induced by Corynebacterium parvum (Cp). Cp, a potent interferon inducer and RES stimulator, has been found to induce both stimulatory and inhibitory effects on NK activity. Cp has been shown to cause profound inhibitory effects on T-cell-mediated immune functions, and many of these negative effects have been attributed to the presence of activated macrophages. The failure to reverse the low levels of NK activity in the spleen of Cp-treated mice, after removal of adherent cells by passage through a nylon column, indicates that Cp either has a direct toxic effect on NK cells or it activates nonadherent suppressor cells, which cannot be physically separated from NK cells by column passage. Some of the biological effects caused by Cp have been found to depend upon T lymphocytes. At the time when depression of NK activity occurred in the spleen of Cp-treated mice, increased levels of cytotoxicity were seen in the peritoneal cavity. The results indicate that in vivo modulation of NK activity by Cp is a complex phenomenon, involving various components of the immune system.

Adherent cells with the ability to inhibit NK activity were found in the spleens of young 8-week-old SJL/J mice. Such suppressor cells were also found in the spleens of other low-NK-reactive strains, but not in the spleen of age-matched high-NK-reactive strains. These suppressor cells had the characteristics of macrophages, since they were plastic- and nylon-adherent, phagocytosed latex and iron particles, and were resistant to treatment with anti-Thy-1.2 serum plus complement. Their inhibitory activity appeared to be due to their production and/or release of soluble suppressive factor(s). Their suppressor activity was similar to that of pyran copolymer-activated macrophages, which were previously shown to be suppressive for cytotoxicity by NK cells. We suggest here that macrophages may be involved in the physiological regulation of NK activity and may contribute to the low NK activity in some strains of mice.

Previous studies suggested that natural killer (NK) cells are involved in the regulation of the growth and differentiation of pluripotent hematopoietic stem cells. To establish whether the effector cells responsible for the in vivo resistance to bone marrow (BM) transplants and the in vitro inhibition of colony-forming units (CFU) may represent identical or overlapping populations, we used a rat system for syngeneic BM transplantation, with and without the transfer of large numbers of peripheral blood large granular lymphocytes (LGLs). BM reconstitution was measured by the in vivo formation of syngeneic CFU in the spleen (CFU-s). Because of the very low frequency of CFU-s in normal rat BM, we fractionated BM cells in Percoll density gradients, which provided a 2- to 5-fold enrichment in CFU-s in the lower-density fractions. Although these fractions contained less than 10% of the total cells, they contained greater than 75% of the CFU-s and allowed for the transfer of significantly fewer donor cells. At the time of BM transplantation, radiation-resistant asialoganglioside GM1-positive LGLs, with high NK activity, accounted for a significant percentage of the lymphoid cells in the irradiated recipient. The in vivo regulatory role of these cells on engraftment was demonstrated by their depletion (by i.v. injection of small amounts of anti-asialo-GM1 antiserum before BM transplantation), which resulted in a significant increase in the number of CFU-s. Conversely, a 50% inhibition in CFU-s was found when CFU-s-enriched BM fractions were preincubated in vitro with LGLs. Additional experiments, involving selective in vivo depletion of NK cells followed by LGL repopulation, directly demonstrated the involvement of LGLs in the regulation and growth of syngeneic pluripotent hematopoietic stem cells. Our results further support the hypothesis that LGLs are involved directly or via humoral factors in the homeostasis and regulation of hematopoietic stem cell growth and differentiation.

A rapid in vivo clearance of tumor cells was found in normal mice following intravenous inoculation of [125l]dUrd-labelled YAC-1 and RBL-5 cultured cell lines derived from lymphomas. The ability of mice to eliminate tumor cells from spleen, liver and lungs during the first 4 h, as evaluated by the recovery of radioactivity in these organs, was found to correlate with the level of natural killer (NK) cell reactivity in their spleen and lungs, as measured simultaneously in vitro in a shortterm 51Cr release assay (CRA). Lower recovery of radioactivity was found in mouse strains with high spontaneous levels of NK activity. The degree of clearance was also found to be age-dependent and older mice of several strains, whose NK activity had declined to low levels, were less effective in eliminating tumor cells. In vivo treatment with interferon and interferon inducers (poly I:C, pyran copolymer, Corynebacterium parvum, murine sarcoma virus) increased the levels of NK activity in the spleen and lungs and also augmented the in vivo clearance of tumor cells from the lungs and liver. Immunopharmacological treatments with antimacrophage agents (silica, iota-carrageenan, Seakem-carrageenan), antineoplastic drugs (dexamethasone, cyclophosphamide, 5-(3,-3′dimethyl-I-triazeno)-imidazole-4-carboxamide, 4-amino-L-D-arabinofuranosyl-2-(IH)-pyrimidone, adriamycin) or irradiation (850 R γ-ray) had comparable effects on both in vitro cytolytic activity and in vivo clearance of tumor cells. When the susceptibility to in vitro and in vivocytotoxicity by several other tumors was examined, the lines with detectable sensitivity to lysis by NK cells were found to be cleared in vivo to a greater degree in a high NK strain (CBA) than in a low NK strain (SJL). In contrast, NK-resistant lines were cleared at approximately the same rate in both strains. However, the actual levels of in vivo clearance and the degree of difference between the strains for the various NK-sensitive lines did not correlate well with their relative sensitivities to lysis in vitro. In the various situations in which the in vivo recovery of a particular NK-sensitive line was studied relative to the levels of NK reactivity in the recipients, the best correlations were seen with clearance from the lungs. In several instances, clearance from the spleen did not correlate well with splenic NK activity. These data indicate that rapid in vivo clearance of radiolabelled NK-sensitive tumor cell lines is appreciably influenced by levels of NK reactivity, but that other factors are probably also involved.