/\98

92nd AACR Immunologic approaches to the treatment and prevention of cancer Keith Knutson

Address Division of Oncology University of Washington 1959 NE Pacific Street Box 356527 HSB BB1321 Seattle WA 98195 USA Email: kknutson@u.washington.edu

IDrugs 2001 4(5):498-500 @ PharmaPress Ltd ISSN 1369-7056

Tumor iJllll11l11ology is advancing at a rapid rate and an increa 'ed Jlll/nuer of investigators are providing illlportant contriuutions to tIJe field. Greater tI/an 250 aus/racts dealing directly with tUlllor immunology were presented at tile 92nd AII/lllat Meeting of the American Association of Cancer Re earcIJ iJl New Orleans, LA, USA. All facets of tumor imr1I1I1IOJogy wrre represented lJy exeelJell1 abstracts addressillg dendritic cells, IlIechanislIlS of effective tumor iJrJmuJlity, imJrJIIlU' defects in callce!" patiwt , Iim70r anligcns, cancer vaccines, and cancer immnnotherapies.

Introduction It was estimated that b tween 10,000 and 15,000 people attended the 2001 Annual Meeting of the American Association of Cancer R search (AACR). The diversified AACR ha' a strong tumor immunology membership, and this group submitted more than 250 abstract for the meting. Mo t of the tumor immunology abstracts could be categorizt::>d using the following descriptors: (i) dendritic cells; (ii) effector cells; (iii) cytokines; (iv) mechanisms of tumor immunity; (v) tumor antigens; and, (vi) immunoth rapy. In general, most of the abstracts were of excellent qua lity, nnd this review highlights some of those presentations that are of pnrticular importance nnd that cover some of the nwre important themes in tumor immunology.

Dendritic cells Tumors evade immune recognition by preventing differentiation of dendritic cells (DCs). P Ch ng and colleagues (Vanderbilt Cancer Center, USA) have identified a protein, myeloid-related protein 8 (MRP8), that may be involved in 0 differentiation and tumor evasion. During normal maturation, DCs downregulate this protein. Immunosuppressive factors secreted by the tumor during culture block both DC differentiation and MRP8 downregulation. Increased MRP8 levcl resulting from mrp8 transfection into DCs prevents differentiation. These results suggest thnt MRP8 may be a target of immunosuppressive factors prod uced by tumor cells.

Fusing DCs with tumor cell lines has b en a common approach to generation of tumor-specific T-cells in vitro and in vivo. However, a limitation with this approach is that tumor cell lines are often difficult to establish and grow ill

vi/roo LM Holmes and colleagucs (C1I1cer Cent l' of Greenville Hospital Syst'm, USA) have developcd il new strategy of deriving instant dendritomns using freshly isolilted primary tumor cells and peripheral blnod mononucleiH cell (PBM )-derived DCs. Tumor cells are labeled with a red dye and Des are labelt::>d with a green dye. Following fusion, the resulting dendritomas are isolated by FA S. Til' d 'ndritoma' were found to express key antigen-presentation mol cules, such as I-1LA class I nnd II. III vitro anlllysis hm·vs thilt dcndritomas generate tUrl1or­specific T-c lis that lyse primary tumor <lnd 'crete interferon (IFN)-y. This strategy may be practical for many tumors in which sufficient tumor cells c<ln b iso!ilted, su h as melanoma.

Effector cells C1audin Rossig and colleagues (Baylor ollege of Medicine, USA) have taken <1 unique approach to g nerating tumor­reactive T-cells ill vitro. A major problem with tumor antigen-specific T-cells m;mipulnted ex vivo is that activity is often low and th' T-cells have a limited lifespan. Dr Rossig explained that Epstein Barr Virus (EBV)-specific T-cells are active in vivo for long periods of tim du to continued antig>n presence. The group transfected EBV-sp cific cytotoxic T-cell lines with <l chimeri receptor protein consisting of the T-cell receptor s-chain fused to the variable domains of th neuroblastoma ganglioside ilntigen, GD2. The effector cell line recognized neuroblastoma cells with IFl 'Y relc<1se and cytolysis. Responsiveness toward autologous B-lymphoblastic cell lines was still maintained, despite neurobla toma responsivene S. Futur studie are to be directed at the c1ini ai, toxicological and immunological effects follOWing infu ion.

The magnitude of the tumor response is carefully controlled by immunoregulalory and suppressor cells. H anaka <1nd colleagu s (Cleveland Clinic Foundation, USA) described a regulatory role of 025+ CD4+ T-cells. Anti-C02 was used to d plete CD25+ CD4+ T-cells from regional tumor­draining lymph node'. Thc remaining lymph node cells had > 3-fold in rea sed antitumor activity compared with lymph node cells without cor depletion. These results suggesl that CD25+ 04+ T-cells exhibit a suppressive regulatory rol in response to tumor, and demonstrate that careful in vitro manipulation of the tumor-spccific T-cells can result in an effective lymphocyte population for potential use in adoptive T-cell therapy.

Cytokines Many cytokines haVing important antitumor attributes have been identified ilnd characterized over the pa t decade. More recently, the importance of interleukin (IL)-18 in the antitumor re ponse has been described. Zdenka Jonak and colleagues (Glaxo mithKline pic, USA) are studying the role of IL-18, and I' ported that administration of IL-18 in both early- and advanced-stage murine tumor models result d in tumor regression and the induction of memory cytotoxic T­

Meeting Report 92nd AACR 4R9

ells. The ilntitumor effects appilrently requir' secondary ilC ation of immune eff~ctor cells, as severe combined II lI10deficient (S 10) mice w r~ unilble to reject tumor f( wing IL-ltl administration. [n particular, they hilve ol rved thilt lL-18 elevates other cytokines, such ilS gral1ulocyte macrophage colony-stimulating factor (,M­

SF), lL-la. ilnd tumor necrosis filctor (TN F)-a., elilborat'd by so-far unchara terized immune eff ctor cells.

Yoshinori Morimoto and colleagues (Okayama University ~·1edicfll School, Jilpiln) evaluated IL-18 responsiv ne's in t n patients with advanced-stage cancer and compared the respon' s to those of ten normal donors. PBMCs from normal blood donors were ilblc to produce significant quantities of IFNy in reponse to IL-18 stimulation, whereas no IFNy production W<lS obst:rved in cancer patients. Further <II alyses showed that cancer patients hOld red uced e. pression of IL-18 re t:ptor ompared t(1 norl1lClI d(1l1ors. Flow cytometric Cll1alysis revealed that ancer patients also had smaller numbers of NK cells that constitutively express IL-l S receptor. Increased understanding of the cytokine defects in can er patients could lead to therapies dir cted at restoring normal cytokine lev Is.

Mechanisms of tumor immunity Dmitry Cabrilovi h and colleagues (1-1 Lee Moffitt Cancer Center, University of South Florida, USA) have previously shown that the number and fun tion of DC' in the peripheral blood WilS dramatically reduced in cancer patients and found that this was linked to an accumulation of immature myeloid cells in the peripheral blood. Upon further analysis, they observed that approximately one-third of immature myelnid cells were immature macrophages and DCs. The addition of these immature cells into PBMC cultures greatly reduced antigen-specific recall responses to tetanus toxoid, as asses ed by prolif ration assay. Furthermore, mature DCs derived in the peripheral blood had increased ability to stimulilte allogenei T- ell responses when the immature myeloid precursors were removed. The authors suggested that causing accumulation of these immature myeloid cells may be il mechanism by which tumor cell, evade d t ction by the immune system.

Tumor antigens The identification of HLA-A and HLA-B pep tides is important for modulating the cytolytic T-cell response to tumors. DC McNeel and colleagues (University of \'\o'ilshington, USA) identified 11 HLA-A2 binding peptides contained within the prostate cancer antigen, prostatic acid phosphatase (PAP). Natural immunity to these peptides was examined in 18 men with prostate cancer and 10 normal male individuals. A significant proportion of both normal individuills and prostate cancer patients recognized three of the I-ILA-A2 epitopes (PAP,..w PAPI12.120 and PAPI3S.Il')· Dr Mc leel explilin d that the presence of naturally occurring immunity to these peptides suggests that they ould be target epitopes for prostate cancer vaccines. Future studie will question the natural presentation of these antigens on prostate cancer cell line .

KD Amos and colleagues (Washington University School of Medicine, USA) investigated seven I-ILA-A2 binding

peptides contained within the brea·t tumor-associated antigen, milmmoglobin A. Mammoglobin A is overexpress'd in > 80% of breast tumor" Two of the HLA-A2 peptides, FMQLlYDSSL and KLLMVLt\·1LA, were fou.nd to elicit CDB T­cells in three out of seven HLA-A2+ normill volunteers. Dr Amos indicated that he is continuing to study the natural response to these peptides in order to determine if mammoglobin could b a potential tOlrget for a breast cancer vaccine. The continued discovery of tumor ilntigen HLA clilss J peptides will greatly facilitate the development of polyantigeni peptide-based cancer va cines. Because CDS T-cells require a concomitant T-celJ helper, intere't in identifying HLA cla's II epitopes of tumor ilntigens has increased. In one important example, Dr Chikilmat u and colleagues (Nationill Cancer Institllte, USA) identified an HLA-DR4 epitop , p5311c>,W of wild type p53. A bulk T-cell line, specific for p5311 c>121' was g nerated from the pt:ripheral blood of a healthy volunteer. The cell line reacted with HLA­DR4+ tumor cell lines, indicating natunl presentation of this epitope. This could be an important addition to peptide­bilsed immunotherapie tnrg ting p53.

Immunotherapy Joachim hultz (Danil-Farb I' ancer Institute In , U A) pres nted a seminar entitled 'On the way to a universal tumor vaccin,'·. Dr huItz and colleagues are inve ·tigating the feasibility of developing a universal tumor vaccine th(1t is effective fit genernting immunity regardles~ of the HLA genotyp of th ancer patient. Th' ideal tumor antigen must have certain characteristics, for example they hav' to be: (i) expre'sed on all or a large majority of tumors regardless of origin; (ii) hidden from toleragenic mechanisms; (iii) necessary for tumor growth and/or survival with no evidence of antigen-loss variants; an I, (iv) have proc ssed epitopes that can b recognized by both CD4 and CDS T-cell'. Dr Schultz and his colleagues hilve potentially identified one such tumor antigen, telomerase, and are investigating the therilp uti potential of UTAC-l, a human telomera e reverse transcriptase-specific peptide. Te!nmerase i a riboprotein enzyme that maintains telomere length stability in all cancer cells. Telomerase confers immortality on canc r cells and immortal germ line cells; telomere shortening is associated with cell senescence. [n preliminary murine studies, immunization against telomerase resulted in the ability of the mic' to reject tumor. Epitopes restricted by both HLA-A2 and I-ILA-A3 have been identified and phase I clinical trials are being planned to ass ss afety of immunizing late-stage, high-risk cancer patients against telomerase.

Copi ShOlnkar and olleagues (Northwest Biotherapeutics Inc, USA) have onducted a phase 1/11 clinical trial evaluating the safety, immunogenicity and clinical response of CaPVax (Northwest Biotherapeutics Inc) in pro ·tate canc r patients. CaPVax for prostate Cilncer consist· of autologous DCs loaded with purified prostate- pecific membrane antigen (PMSA). Vaccination was safe with negligible adverse reactions. A total of 91% (10/11) of patients had in reased PMSA-specific antibodies following va cination, while 80% dev loped vilrying T-cell responses to the vaccine. Although patient number were too small for statistical conclusion, 80';!0 (four of fiv) patients that generated high levels of immunity to the vaccine had

"00 /Drugs 2001 Vol 4 No 5

tabilized di 'ease ompared to 25'}\, of the pati nts (one of four) that generilt d low levels of c lIular immunity. Dr Shankilr indicated that Northwest Biotherilpeuti s is interested in taking this immunotherapy to a phase III clinical trial to further ascertain clinical efficacy in the therapeutic setting. ­

An important is ue for tumor vaccinologists is the threat of genernting Jntigen-Ioss variants following immwlization, which would circumvent the potential protective or urative -'ff cts of ;'I vaccine. X Krlng ;md colleagues (ImCione Systems Inc, USA) have designed and constructed a nove) himeric protein, hTRPx3, consisting of the human TRP-1, TRP-2 and tyrosinase proteins. hTRPx3 W;'lS tested in a preclinical mod I for the generati 111 of immune respons s. Both antibody and T­cell I' spon es were elicited in mic following immunization. The vaccine was effective with many adjuvant preparations and ould protect mice from tumor challenge.

SUppJ"ssion of antigen-presenting cell (CtlVlty in ancer pntients is cummon. Sylvia Kiertscher ;'Ind colleagues (University of California Los Angele' hool of Medicine, USA) hypothesized that increasing the numbers of functional circulClting antigen-presenting cells in cancer p<ltients may impl'llve endogenous antitumor immunity. The u~e of GM- F and IL-4 in combination resulted in generi1tinn of DCs in vitro. A pha e I dose-escalation study, i1dministering both GM-CSF and IL-4, was conducted in cancer patients with treatment-resistant disease and patients who refused treatment. A total of 21 patients were treated for 7 days in six cohorts, with vMying doses of IL-4 (0.5 to 6 mg/kg/day) and a constant dose of GM-CSF (2.5 mg/kg/day). Following treatment, it was observed that the DCl population (Lin-, DR+, CDllc+) increased substantially in the peripheral blood from 8 x 10' cells/ml to 8.7 x 10" cells/ ml. This treatment abo resulted in the increased expression of HLA-DR on the D 1. Thes results suggest that cytokine treatment may improve tumor immune responses dir ctly ill vivo through enhanced antigen presentation. Although regression of established tumor WZlS

limited with this approach, cytokine therilpy coupled with vaccinntion may improve tumor-specific immunity.

In recent years, th use of novel "ntibody constructs to passively treat mnlignilncy has incr ased. Oliver Pre sand colleagues (University of Washington, USA) reported results of a clinical trial eVilluating th fficacy of radioiodinated anti-CD20 ('Jl1 anti-CD20; NeoRx) for the treatment of B-cell

lymphomas. CD20 is expressed on > 95'1';. of B-cell lymphomas. In patients with newly diagnosed l3-cell lymphoma, 63% responded to nil anti-CD20 therapy with a complete respon e, while 34%. responded with a partial r sponse. A tot,,1 of 34% of p<ltients with relapse disease had a complete response and 37"i\. had a p"rtial I' 'sponse. Piltients with refrilctory di 'eilse responded less well, with 17'1., "chieving iI complete response and 48% a parti,,1 response.

Summary The field ()f tumor immunology is as diverse as the maligmncies themselves. Currently, immunological strategies will likely h"ve to be tailored to the type of malignancy, as well as the characteristics of the p<ltient, until a universal treatment or va cinntion m -thod is esti1blished. [s telonprnse the ilnswer? Can er vaccines will likely be more effectiv for th pI' vention of relapse or possibly even the de 1I0VO devclopment of new tumors. Stl'ilkgics for effcctiv' prophylactic v<1ccination 3rC emerging and in lude peptide-, protein- and DNA-based v"ccines targeting tumor­ilssociated antigens.

Although it is now understood thnt cancer piltients demonstrate immunosuppression, thi . may be overcome by using vaccine with the appropriate cytokines, such as GM­CSF and IL-4 (ie, combination npproaches). Fm existing malignancie', a wealth of cellulilr- and ilntibody-b"sed ·trategies exists which can be individually tailored to a spe ific tumor. An important example of this is the use of 1J'l anti- D20 antibody to tilrget newly diagnosed B­lymphom<1·. Active cellular-based str<ltegies, rather thiln passive ilntibody-based striltegies, such as adoptive T-cell or DC therapy may likely be more effective for solid tumors, unless the tumor burden is substantially reduced. The use of cytokines to modulate and sustain the immune response, a' well as correct immune defects in cancer piltients, is becoming an increasingly important issue.

Finally, th volution of antigen-loss variants, following effective vaccination or th rap)', point~ to th' need of polyantigen vaccines and therapies. ImCione' development of a chimeric multi-antigen protein va cine to target melanoma provide' an exc lIent example. The continued identification and haracterization of tumor-a 'so iated antigens paves the ,·vay for development of poly-ilntigen stra tegies.