The Biology Behind

Natural Killer ^ Based Immunotherapy 55 CommentaryonMotohashietal.,p.6079 HansJ.J. van der Vliet,1, 2 Steven P. Balk,1and Mark A. Exley1

In this issue of Clinical Cancer Research, Motohashi et al. (1) lymphoma, sarcoma, melanoma, and lung carcinoma, suggest- present data on the first clinical trial using adoptive transfer ing broad clinical applicability(reviewed in ref. 6). On pre- of invariant natural killer T (iNKT) cells in patients with sentation of a-GalCer byCD1d-expressing -presenting advanced or recurrent non–small cell lung cancer. This study cells, iNKT are activated and rapidlyproduce large amounts of follows up on several other recentlydone clinical trials targeting predominantlyIFN- g but also IL-4 that cumulativelyresult in iNKT that all aim to induce the same robust antitumor effects the activation of NK cells, B cells, CD4+ and CD8+ T cells, in humans as previouslyobserved in murine tumor models. and the antigen-presenting cells with which theyinteract. iNKT constitute an evolutionary conserved T Through the production of IFN-g, as well as through direct lineage that displays an extremely restricted T-cell antigen CD1d signaling (7), iNKT enhance maturation receptor repertoire with limited (V a24-Ja18 and IL-12 production. It is this cascade of events that is even- in human and Va14-Ja18 in mouse). iNKT can recognize tuallybelieved to result in the development of a Th1-biased certain glycolipid , which may include the endogenous proinflammatoryantitumor immune response that is pivotal isoglobotrihexosylceramide and the synthetic a-galactosylcer- for the antimetastatic activityof a-GalCer bypromoting the amide (a-GalCer), in the context of the monomorphic generation of a long-lasting antitumor effector cell population nonclassic MHC class I–like antigen-presenting molecule and byinhibiting tumor angiogenesis (8, 9). CD1d. One of the features that distinguish iNKT from CD1d is constitutivelyexpressed bydendritic cells and other conventional T cells is their capacityto rapidlyand in some antigen-presenting cells at low levels bycertain other tissues cases simultaneouslyproduce large amounts of both T helper and can be up-regulated in response to stress (2, 8). The (Th) 1 proinflammatory(promoting cellular immune possible contribution of tumor cell CD1d expression remains responses) and Th2-type anti-inflammatory (promoting certain unclear. iNKT can mediate antitumor immunitythrough the antibodyresponses) cytokinesupon triggering, therebymedi- activation of other immune cells, in which case tumor ating the regulation of a varietyof immune responses, regression would be independent of tumor cell CD1d including antitumor immune responses (2). expression. Direct lysis of CD1d-expressing tumor cells has iNKT playa physiologicrole in tumor immunosurveillance been reported (10), however, although some tumors might against carcinogen-induced tumors (3) and are required for escape direct lysis by iNKT through shedding of neutral the antitumor effects of low-dose interleukin (IL)-12 treat- glycolipids (11) and/or their absence of costimulatory signals ment (4) through their capacityfor Th1 cytokineproduction that could lead to altered iNKT activation and consequent (Fig. 1). Manystudies designed to evaluate the antitumor suppression of antitumor responses (2, 8). activities of iNKT have used the model and highlyspecific Clinical evaluation of the antitumor activityof iNKT began glycolipid antigen a-GalCer (KRN7000) that was originally when i.v. administration of a-GalCer was evaluated in a phase isolated from the marine sponge Agelas mauritianus in a screen 1 clinical trial in patients with solid tumors. Although no for novel antitumor agents bythe KIRIN pharmaceutical clinical responses were observed, this studyshowed that a- company(Gunma, Japan; ref. 5). Antitumor effects of a-GalCer GalCer was well tolerated and indicated the importance of a were observed in a varietyof tumor metastasis models of liver, sizeable pool of iNKT in patients treated with a-GalCer, as signs lung, and lymph nodes, including colon carcinoma, T-cell of immune activation onlyoccurred in the smaller subset of patients with relativelynormal iNKT numbers (12). Simultaneously, studies showed quantitative and qualitative defects in the iNKT pool in various types of cancer, including colon cancer, lung cancer, breast cancer, melanoma, head and 1 Authors’ Affiliations: Cancer Biology Program, Division of Hematology and neck squamous cell carcinoma, prostate cancer, myelodys- Oncology, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, Massachusetts and 2Department of Internal Medicine, Vrije Universiteit plastic syndromes, and progressive malignant melanoma, Medisch Centrum, Amsterdam, the Netherlands although not in glioma (reviewed in ref. 6). Murine experi- Received 6/8/06; accepted 6/13/06. ments indicated that the antitumor activityof a-GalCer could Grant support: Netherlands Organization for Scientific Research TALENT grant be enhanced when a-GalCer was loaded onto dendritic cells and grant 920-03-142, Dutch Cancer Society academic grant, NIH grants R01 (13–15), this has thus far been evaluated in three clinical DK066917 and R01AI42955, Harvard Center for Human Cell Therapy pilot grant, Hershey Family Prostate Cancer Research Fund, and Dana-Farber/Harvard Cancer studies. Indeed, therapywith a-GalCer-pulsed dendritic cells Center Skin Cancer Specialized Program of Research Excellence grant P50 seemed more potent and induced inflammatorytumor CA93683. responses, tumor necrosis, and decreases in tumor markers in Requests for reprints: Mark A. Exley, Beth Israel Deaconess Medical Center, several patients as well as expansion of iNKT over several weeks NRB 1030L, 330 Brookline Avenue, Boston, MA 02115. Phone: 617-667-0982; E-mail: [email protected]. to a few months and an increase in adaptive T-cell F 2006 American Association for Cancer Research. (16–18). Again, immunologic responses were most prominent doi:10.1158/1078-0432.CCR-06-1392 in patients with relativelyhigh numbers of circulating iNKT

www.aacrjournals.org 5921 Clin Cancer Res 2006;12(20) October 15, 2006 Downloaded from clincancerres.aacrjournals.org on September 29, 2021. © 2006 American Association for Cancer Research. The Biology Behind

Fig.1. Simplified model for recognition modes by iNKT.To the left, normal antigen-presenting DC enhanceTh1 antitumor responses through a MHC class 2-like interaction also involving co-stimulation (red hatched to left) and DC Th1cytokines such as IL-12 (red hatched to right).Tothe right, presentation by tumor cells lacking co-stimulation and secreting suppressive such asTGF-h (blue hatched to right) results in iNKT activation toward Th2 response (IL-4, IL-13) and consequent suppression of antitumor imunity. An intermediate effect might be expected if immature DC or defective intratumoral DC were the predominant presenting cells. Other variables include potential glycolipid antigens from the tumor presented directly by the tumor and/or indirectly via DC to the iNKT, the predominance of certain iNKTsubsets, and the presence of other immunoregulatory cells.

(17) and in vivo iNKT responses were strongest using mature ena, which is a concern when transferring cells that seem to dendritic cells (18). regulate self-tolerance. Motohashi et al. selected patients for The quantitative defects in iNKT numbers that are observed inclusion in their studywith relativelyhigh amounts of in manycancer patients are accompanied byqualitative circulating iNKT. Although this is reasonable for toxicity defects in the residual iNKT population, including defective analysis, one might expect that patients with more deficient IFN-g production (19), which therebyhamper antitumor iNKT populations would benefit most from administration of effects of the iNKT ligand a-GalCer. Using this line of iNKT and that this patient group should not be excluded in thinking, beneficial clinical results maybe expected from future trials, which ideallywould also be of larger scale immunotherapeutic strategies directed at expansion and allowing evaluation of clinical responses. activation of Th1-polarized iNKT in cancer patients. This is Clearly, the article of Motohashi et al. sets the stage for further supported bymodels in which the antitumor activity further studies that will focus on optimization of this of adoptivelytransferred iNKT was shown in murine lung interesting type of immunotherapy. We believe that apart from cancer and melanoma (14, 15) and byaccumulating evidence attempts to use more purified populations of iNKT, future in man indicating that numerical defects in the size of the studies should additionallyfocus on a comparison of the effects iNKT pool are indeed of clinical significance, as theyare of different iNKT subsets, as these have been shown to be associated with poor prognosis in patients with neuroblasto- functionallydistinct. CD4 + iNKT produce both Th1- and À À ma and colon cancer (20, 21). Th2-type cytokines, whereas CD4 CD8 double-negative In this issue of Clinical Cancer Research, Motohashi et al. iNKT, and especiallythe rarest subset, CD8 + iNKT, produce describe the results of a clinical phase 1 studyin which six Th1-type cytokines (22–24). As antitumor immune responses patients with advanced or recurrent non–small cell lung cancer have traditionallybeen reported to benefit most from Th1-type received autologous in vitro –expanded iNKT that were cul- immune responses, one would expect that these different iNKT tured from leukapheresed peripheral (1). Although subsets would differentiallyaffect antitumor immune responses treatment did not result in clinical responses, it was well and this was indeed shown in recent preclinical models in tolerated and resulted in some interesting immunologic which double-negative iNKT, but not CD4+ iNKT, from murine phenomena, although some cannot with certaintybe tracked liver facilitated rejection of sarcoma and melanoma (25). As back to iNKT as the purityof the infused iNKT product was iNKT from different tissues have different antitumor potential relativelylow (0.3-25%), and included manyclassic T and NK (i.e., liver > spleen > thymus; ref. 25), identification of factors cells. In addition, cells were transferred while still activated. responsible for these differences could be of benefit in the This might encourage an initial response, but risks rapid selection and generation of iNKT that would be expected to be clearing, as is the fate of classic activated T cells, hence the very most beneficial in clinical studies. In this regard, it is large doses used in trials of the latter cells. As desired, the study noteworthythat the iNKT transferred byMotohashi et al. were resulted in increased numbers of circulating iNKT in several largelythe desired double-negative iNKT. A relativelyhigh patients and was not accompanied byautoimmune phenom- frequencyof CD8 + iNKT was observed in several individuals,

Clin Cancer Res 2006;12(20) October 15, 2006 5922 www.aacrjournals.org Downloaded from clincancerres.aacrjournals.org on September 29, 2021. © 2006 American Association for Cancer Research. NKT Cell ^ Based Cancer Immunotherapy

although it remains uncertain whether these cells represented imagine that, to allow the full induction of protective immunity true CD8ah+ or CD8aa+ iNKT (i.e., activated double-negative byiNKT, iNKT therapymight need to be combined with iNKT). Other approaches that seem worthyof future clinical CD4+CD25+ regulatoryT-cell depletion, for example, by evaluation include the combination of iNKT transfer followed denileukin diftitox (28). by a-GalCer treatment (ideallyin the context of dendritic The insight into the biologyof iNKT and immunoregulatory cells that can simultaneouslybe loaded with tumor-associated T cells in general has expanded rapidlyover recent years.iNKT antigens), iNKT therapyafter chemotherapypretreatment as therapybyitself might not be sufficient to promote long-lasting this seems to sensitize tumor cells to iNKT-mediated cytotox- immune responses, but it has the potential to substantially icity(26), and direct modulation of CD1d on antigen- stimulate the development of conventional long-lasting anti- presenting cells (e.g., byanti-CD1d monoclonal ) tumor immune responses when combined with, for example, as this has been shown to trigger the production of the dendritic cell–based vaccination. Th1-promoting IL-12 (6). Interestingly, CD4+CD25+ regulatoryT cells, which constitute another subset of immu- noregulatoryT cells (reviewed in ref 1), activelydown-regulate Acknowledgments immune responses, including antitumor responses and iNKT + + responses (27). As this population of CD4 CD25 regulatory We apologize to all our many colleagues whose work had to be omitted or cited T cells is frequentlyexpanded in cancer patients, one can indirectly through reviews because of space and reference limitations.

References 1. Motohashi S, Ishikawa A, Ishikawa E, et al. A phase I 11. Sriram V, Cho S, Li P, et al. Inhibition of glycolipid 19. Tahir SM, Cheng O, Shaulov A, et al. Loss of IFN-g study of in vitro expanded NKT cells in patients with shedding rescues recognition of a CD1+ Tcell lympho- production by invariant NK Tcells in advanced cancer. advanced and recurrent non-small cell lung cancer. ma by natural killerT (NKT) cells. Proc Natl Acad Sci J Immunol 2001;167:4046 ^ 50. Clin Cancer Res 2006;12:6079 ^ 86. USA 2002;99:8197 ^ 202. 20. Metelitsa LS,Wu HW,Wang H, et al. Natural killerT 2. Kronenberg M, Rudensky A. Regulation of immunity 12. Giaccone G, Punt CJ, Ando Y, et al. A phase I study cells infiltrate neuroblastomas expressing the chemo- by self-reactiveTcells. Nature 2005;435:598 ^ 604. of the natural killer T-cell ligand a-galactosylceramide kine CCL2. J Exp Med 2004;199:1213^ 21. 3. Smyth MJ,Thia KY,Street SE, et al. Differential tumor (KRN7000) in patients with solid tumors. Clin Cancer 21.TachibanaT, Onodera H,TsuruyamaT, et al. Increased surveillance by natural killer (NK) and NKTcells. J Exp Res 2002;8:3702^ 9. intratumor Va24-positive natural killerT cells: a prog- Med 2000;191:661 ^ 8. 13. Toura I, Kawano T, Akutsu Y, Nakayama T, Ochiai T, nostic factor for primary colorectal carcinomas. Clin 4. Takeda K, Hayakawa Y, Atsuta M, et al. Relative Taniguchi M. Inhibition of experimental tumor metas- Cancer Res 2005;11:7322^ 7. contribution of NK and NKTcells to the anti-metastatic tasis by dendritic cells pulsed with a-galactosylcera- 22. Lee PT, Benlagha K,Teyton L, Bendelac A. Distinct activities of IL-12. Int Immunol 2000;12:909 ^ 14. mide. J Immunol 1999;163:2387^91. functional lineages of humanVa24 natural killerTcells. 5. Natori T, Morita M, Akimoto K, Koezuka Y. 14. AkutsuY, NakayamaT, Harada M, et al. Expansion of JExpMed2002;195:637^41. Agelasphins, novel antitumor and immunostimu- lung Va14 NKTcells by administration of a-galactosyl- 23. Gumperz JE, Miyake S, Yamamura T, Brenner MB. latory cerebrosides from the marine sponge Agelas ceramide pulsed dendritic cells. Jpn J Cancer Res Functionally distinct subsets of CD1d-restricted natu- mauritianus. Tetrahedron 1994;50:2771 ^ 84. 2002;93:397 ^ 403. ral killer T cells revealed by CD1d tetramer staining. 6. van derVliet HJ, Molling JW, Blomberg BM, et al. The 15. Shin T, Nakayama T, Akutsu Y, et al. Inhibition J Exp Med 2002;195:625^ 36. immunoregulatory role of CD1d-restricted natural killer of tumor metastasis by adoptive transfer of IL-12- 24. Takahashi T, Chiba S, Nieda M, et al. Cutting edge: Tcells in disease. Clin Immunol 2004;112:8 ^ 23. activated Va14 NKT cells. Int J Cancer 2001;91: analysis of human Va24+CD8 + NK T cells activated 7. YueSC,ShaulovA,WangR,BalkSP,ExleyMA.CD1d 523 ^ 8. by a-galactosylceramide-pulsed monocyte-derived ligation on human monocytes directly signals rapid 16. Nieda M, Okai M, Tazbirkova A, et al. Therapeutic dendritic cells. J Immunol 2002;158:3140 ^ 4. NF-nB activation and production of bioactive IL-12. activation of Va24+Vh11NKT cells in human subjects 25. CroweNY,CoquetJM,BerzinsSP,etal.Differential Proc Natl Acad Sci USA 2005;102:11811^ 6. results in highly coordinated secondary activation antitumor immunity mediated by NKT cell subsets 8. Godfrey DI, Kronenberg M. Going both ways: im- of acquired and innate immunity. Blood 2004;103: in vivo. J Exp Med 2005;202:1279 ^ 88. mune regulation via CD1d-dependent NKTcells. J Clin 383^9. 26. Mattarollo SR, KennaT, Nieda M, Nicol AJ. Chemo- Invest 2004;114:1379 ^ 88. 17. Ishikawa A, Motohashi S, Ishikawa E, et al. A phase I therapy pretreatment sensitizes solid tumor-derived + 9. HayakawaY,Takeda K,Yagita H, et al. IFN- g-mediated study of a-galactosylceramide (KRN7000)-pulsed cell lines to Va24 NKT cell-mediated . inhibition of tumor angiogenesis by natural killer dendritic cells in patients with advanced and recurrent Int J Cancer 2006;119:1630 ^ 7. T-cell ligand, a-galactosylceramide. Blood 2002;100: non-small cell lung cancer. Clin Cancer Res 2005;11: 27. AzumaT,TakahashiT, KunisatoA, KitamuraT, Hirai H. 172 8 ^ 3 3 . 1910 ^ 7. Human CD4 +CD25 + regulatory T cells suppress NKT 10. Metelitsa LS, Naidenko OV, Kant A, et al. Human 18. Chang DH, Osman K, Connolly J, et al. Sustained cell functions. Cancer Res 2003;63:4516^ 20. NKT cells mediate antitumor cytotoxicity directly by expansion of NKT cells and antigen-specific T cells 28. Dannull J, Su Z, Rizzieri D, et al. Enhancement of recognizing target cell CD1d with bound ligand or in- after injection of a-galactosylceramide loaded mature vaccine-mediated antitumor immunity in cancer directly by producing IL-2 to activate NK cells. dendritic cells in cancer patients. J Exp Med 2005; patients after depletion of regulatory T cells. J Clin J Immunol 2001;167:3114^ 22. 201:1503 ^ 17. Invest 2005;115:3623 ^ 33.

www.aacrjournals.org 5923 Clin Cancer Res 2006;12(20) October 15, 2006 Downloaded from clincancerres.aacrjournals.org on September 29, 2021. © 2006 American Association for Cancer Research. Natural Killer T Cell−Based Cancer Immunotherapy

Hans J.J. van der Vliet, Steven P. Balk and Mark A. Exley

Clin Cancer Res 2006;12:5921-5923.

Updated version Access the most recent version of this article at: http://clincancerres.aacrjournals.org/content/12/20/5921

Cited articles This article cites 28 articles, 18 of which you can access for free at: http://clincancerres.aacrjournals.org/content/12/20/5921.full#ref-list-1

Citing articles This article has been cited by 4 HighWire-hosted articles. Access the articles at: http://clincancerres.aacrjournals.org/content/12/20/5921.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://clincancerres.aacrjournals.org/content/12/20/5921. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.

Downloaded from clincancerres.aacrjournals.org on September 29, 2021. © 2006 American Association for Cancer Research.