Introduction

HER2 as an Important Tumor Antigen The HER2/neu oncogenic protein is a well-defined cancet-related immunogenic protein (or tumor antigen) that is expressed by multiple tumor types. Biologically, HER2 expression in tumors is associated with more aggressive disease. Patients whose tumors overexpress HER2 have a higher risk of relapse and shortened overall survival compared to patients whose rumors do not over­express HER2. t These observations provided the basic rationale for the development of anticancer strategies that specifically target the HER2 protein. In the past 10 years, several efforts have been made to selectively target tumo~ that overexpress HER2, including the develop­ment of HER2-specific therapeutic antibodies and HER2-directed vaccinations. Thi~ ~mide summarizes the most recent data available from clinical efforts to develop HER2 vaccinations that will elicit antirumor immune response.

Theoretical Therapeutic Potential for HER2 Vaccination Generating an active immune response directed against the HER2 protein has several potential clinical advan­tages. Vaccination, if effective, would stimulate immunologic memory and might reduce the.risk of relapse after standard therapies, such as surgery and radi­ation. Moreover, an immunization strategy that selected appropriate epitopes could elicit production of durable levels of functional antibodies. Immunologically, portions of this transmembrane protein are likely CO be available to the class IIIl antigen-processing pathways, thus stimulating a T-helper cell or a cytotoxic T-cell (CTL) immune response (Figure 1). The HER2 protein. therefore, is a good vaccine target (rom both a biologic and immunologic standpoint.

Perhaps more compelhng evidence that the HER2 protein might be a reasonable vaccine candidate is the observ;l6on that some patients with HER2-overex­pressing tumors have low-level preexistent immunity

Dr. Disis is a member of the Speaker's Bureau of lmmunex. This article includes discussion of mvesti!!ational and/or unlabeled uses of dru~s: tncludinl" the use of HER2-()etHide -vaccines.

Mary 1. Disis; MD .__ ~ Associate Professor

Division of Oncology UniveT.lity of Washington

Associate Member Fred Hutchinson Cancer Research Center

Seattle, WA

directed against the protein. Patients with, HER2 protein-overexpressing breast, ovarian, colon, prostate, and non-smaH-cell lung cancers have been shown to have antibodies and/or T cells recognizing HER2 (Table 1). However, no matter what the tumor type, endoge­nous HER2-specific immune responses are found in only a minority of patients whose tumors overexpress HER2; the immunity, if detectable, is low in magnitude compared to an immune response against a foreign anti­gen such as tetanus or cytomegalovirus.2-6 These obser­vations have led to the development of vaccine strate­gies designed to boost HER2 immunity in a majority of patients. HERl is a nonmurated self-protein; therefore, vaccines must be developed based on immunologic prin­ciples focused on circumventing tolerance, a ptimary mechanism of tumor immune escape.

Design and Selection of HER2 Peptide-Based Vaccines

HER2-specific vaccines have been rested in human clin­ical trials (Table 2).7- 12 The majority of currently published clinical triab ofHER2 vaccines has focused on immunizing patients with fragments of the HER2 protein or peptide;; designed to stimulate a specific immune response.

The-use of synth~tic peptides as cancer vaccines offers practical advantages, such as ease of construction and production, chemical sta.bility, and a lack of infectious or oncogenic porential. 13 Peptides might also allow better

. manipulation of the immune response by using epitopes designed to stimulate particular subsets of T cells, ie CD4-positive or CD8-positive T cells. Most importand)'. peptide vaccines appear [Q be effective in generating immune responses ro se1f-proteins.l4 T cells might be tolerant to the dominant peptide epitopes of self-proteins but might respond to 5ubdominant peptide epitopes.l5,16

A~ many newly defined tumor antigens such as HER2 are self-proteins, peptide immunization might playa key role in the ability to elicit an immune response to these antigens. 17

m

might allow more efficient presen­tation of peptide epitopes than standard adjuvants such as IFA. Six patients with HERZ-overex­pressing cancers received 6 monthly vaccinations with 500 pg of p369-377 admixed with 100 Il8 of GM-CSE The patients had stage III/IV breast or ovarian cancer. Immune responses to p369-377 were examined using an inter­feron-r enzyme-linked immuno­spot assay, which enabled enumer­ation of functional peptide-specific CTLs. Prior to vaccination, the med ian precursor frequen c y, defined as precursors/106 periph­eral blood mononuclear cells, CTLs recognizing p369-377 were not detectable. Following vacci­nation, HER2 peptide- specific precursors developed to p369-377 in just 2 of 4 evaluable subjects. The responses w~re short-lived and not detectable at 5 months after

The HER2 protein expreo,cd on the sllrface ofa cancer cell is composed of.n £CD and an lCD. ImIllceUuIar proteiN are proc<:ssOO through me the final vaccination. Immune MHC cl.... I pathway and stimulate a CD8-positive en response thot can result in di""" cell deam. Eltt(accllu[ar proteins ale processed throuRh the MHC class II pathway, uoually aftc, being phagocyti.ed by an APC, and 'timul.te a C04-"",itive T-helper '<»lOMe. CD4-po!itive competence was evident as T cells differentiate into a Th2 phenOtype, secreting ,pec.;nc cytokir= .uch as ll.,.IO after stimulation, or a Th I phenotype secreting cyookme. patients had detectable T-cell ",ch as IL-2, A Th2 CD4-"",itive T-cell response will .ugment B-ceU immunity and the producnon of an'ibod ieo, and a Th J C04-p""itive T­cell respome will augmen' ,he prohferarion of COB-positive en. Antibody immunity al50 has Q potentially therapeutic effect .. rhe anribody responses to tetanus toxoid and c;\I\ bind <he HER2 ECD and inhibit signaling through the growth factor receptor. ;nfluenza~ These findings bring to Abbreviation,,; APe = .nrigen-presenting cell; CTL ~ cy'otoxic T cell; ECD p exr;rncellular dom.in; leD = intracellular domain; IFN = interferon; IL ~ \nterleukll\; MHC - major histocomp.tabi\iry compI.,;

Stimulating CDS-Positive Cytotoxic Lymphocytes With Human Leukocyte Antigen-A2-Restricted Peptides

The CTL has been considered the primary effector cell of the immune system capable of eliciting an antitumor response. CTLs can directly lyse tumor cells expressing specific peptide-major histocompatibility complex struc­tures on their surface {Figure O. Several initial studies immunizing patients against HERl focused on stimulating elL by vaccination with a HERZ-derived human leuko­cyte antigen (HLA) -A2-restricted peptide, p369-377 .18

The first published study of a HER2 vaccine immunized HLA-AZ-positive patients with metastatic HER2-overex­pressing breast, ovarian, or colorectal carcinomas every 3 weeks with 1 mg of p369-377 admixed in incomplete Freund's adjuvant (lFA).7 Peptide-specific ens were isolated and expanded from the peripheral blood of patients after 2 or 4 immunizations. These CTLs could lyse HLA-matched peptide-pulsed target cells but could not lyse HLA-matched tumOts exptessing the HER2 protein.

A SImilar study was performed immunizing patients with p369-377 using granulocyte-macrophage colony-stimu­lating (actor (GM-CSF) as an adjuvanr.8 GM-CSF is a recruitment and rnaturation factor for skin dendri tic cells (DCs), or Langerhans cells (LCs) and, theoretically,

light the potential problems with using a single HLA-A2-binding

peptide for vaccination; the immunity generated is low­level and short-lived.

Stimulating CD4-Positive T-Helper Lymphocytes With Helper Peptides

T-helper celis are involved in the augmentation of both B-cell- and T-celt-mediated immunity. Studies immu­nizing patients with peptides designed to stimulate a CD4-positive T-helper response have been more suc­cessful in generating HER2-specific T-cell responses that are both detectable and durable. As an example, Knutson and colleagues immunized breast and ovarian cancer patients with putative T-helper epitopes of HERl that had HLA-A2-binding motifs ofHER2 embedded in the natural sequence. Thus, both CD4-positive and CD8-positive-specific epitopes were encompassed in the same vaccine. In this trial, 19 patients with HER2-over­expressing cancers ren:ived a preparatory vaccine con­taining putative HER2 T helper peprides.9 Patients developed both HER2-specific CD4-positive and CD8­positive T-cell responses. The level of HER2 immunity was similar to vital and tetanus immunity. In 3ddition, the peptide-specific T cells were capable of lysing tumor cells. The responses were 10I'lg-lived and detectable for> 1 year after the final vaccination in selected pattents.

Stimulating an effective T-helper response is a way to

boost antigen-specific immunity as CD4-posirive Teells generate the eytokine environment required to support an evolving immune response. In an additional study vaccinating patients with HERZ T-helper derived peptiJes, patients with advanced-stage. HERZ-overex­pressing breast, ovarian, and non-smaIl-cell lung cancers were enrolled. Thirty-eight patients finished the planned course of 6 immunizations.10 Patients received 500 ~g of each peptide admixed in GM-CSEI9 Over 90% of patients developed Tcen immunity to HERZ peptides and> 60% to a HER2 protein domain. Thus, immuniza­tion with peptides resulted in the generation of T cells that could respond to HERZ processed by antigen­presenting cells (APCs). Furthermore, at i-year follow­up, immunity to the HERl protein persisted in over a third of patients. Immunity elicited by active immuniza­tion with CD4-positive T-helper epitopes was durable.

Inducing Epitope Spreading

An additional finding of this study was that epitope spreading was observed in the majority of patients and significantly correlated with the generation of HERZ protein-specific T-cell immunity. Epitope, or determi­nant, spreading is a phenomenon first described in autoimmune disease.20 It represents the generation of an immune response to a particular portion of an immuno­genic protein and then the natural spread of that immu­nity to other areas of the protein or even to other anti, gens presenr in the environment.

Epitope spreading was reported recent!y in a vaccine trial immunizing breasr and ovarian cancer patients with autologous DCs pulsed with polymorphic epithelial mucin or HER2 peptides. IJ In this trial, 10 patients were immunized. Half the patients developed CDS-positive T-cell precursors in response ro their immunizing peptides. Moreover, some patients developed new immunity to other tumor antigens expressed in their cancers, such as carcinoembryonic antigen and melanoma-associated antigen-J.

How docs epitope spreading develop? Theoretically, the initial immune response can create a microenvironment at the site of the tumor that enhances endogenous local immunity.ZI Clinical studies demonstrating epitope spreading have relied on the use of DCs as the APCs initiating the immune response. Most clinical trials of cancer vaccines focus on the detection of a newly gener­ated immune response or the magnitude of the antigen­specific immune response elicited after active immunha­tion. However, the detection of epitope spreading that indicates an immune microenvironment capable of producing an endogenous polyclonal immune response might be an endpoint that could potentially reflect an improved clinical outcome. Despite rhe generation of detectable immunity againsr the sdf-protein HER2 and the development of epitope spreading, none of the patients in the studies described above developed any

Tabre 1

Breast3f Antibody

t--B"",'r",,:.. e_as_'~_'t-'-t-''----! ~ll Colon" .Antibody

CU{I;105- ' T c.ell

Prostate6 Antibody

3/45 (7%)

5/450i%) 8/57 (14%)

5/22.(23%)

• Early.ra,,"" t Late .tage

evidence of autoimmunity directed against tissues exp~essing basal levels of HER2 such as skin, liver, and digeStive tract ePithelium. AClministration of HERZ vaccines was safe and well tolerated.

Using Cytokines to Expand and Activate Antigen-Presenting Cells

Recent studtes have evaluated vaccine strategies focused on maximizing the role of the most efficient APC, DC, or skin LC in eliciting effective immunity to self. One such strategy is to use cyeok-ines involved in DC produc­tion and maturation as vaccine adjuvants. Flt3 ligand (FL) is a cytokine that, when administered systemically, can increase numbers of circulating DCs > 40-fold.22

Furthermore, activation of DCs in vivo by FL has been shown to be an effective way of circumventing tolerance during active immunization in animal mooels.z3 Based on data generated in rodent models, 10 patients with HER2-overexpressing cancers were enrolled to reCelve a

Table 2

Zaks and Rosenberg?

HLA-A2 CTL IFA

Breast Ovarian

Colo-rectal 4

Knutson et alB HLA-A2 en GM,CSF Breast

Ova<ian 6

Knutson et al9 Thelpe< GM-CSF Breast

Ovarian 19

Disis et apo

Brossart et all!

T helper

HlA-A2 CTL

GM-CSF

Dendritic cells

Breas( Ovarian 45 NSCLC

,,-----,

Breas( Ovarian

10

Disis et a[l}. Thelper GM,CSF

Flt3 ligand Breast

Ovarian 10

Abbreviations, CTL =cytoroxic T edl, GM-CSf = granulocyte-macrophage colony-stimulating faemr; HLA ~ human leukocyte antigen; IFA = Incomplete Fteunci', adJuvam; NSCLC = nOl\-8mall· ee II·hIn g cancCT

HERl peptide-based vaccine targeting the intracellular domain (lCD) of the HER2 protein. 12 The peptides in the vaccine were the same as those used in one of the arms of the trial described above admixed widl GM-CSF alone as an adjuvant. All patients received FL 20 JIdkg/day s.c. for 14 days. to Five patients received the HER2 peptide-based vaccine alone intradermally midpoint in 1 FL cycle, and 5 patients received the vaccine admixed with 150 IJg ofGM-CSF intradermally midpoint in the FL cycle. All patients showed a substan­tial boost in the frequency of B cells and of T cells responding to both the HER2 ECD peptides and the ICD protein. The small sample size of each group, however, did not aHow astatistlcally significant compar­ison of immune responses between the FL-alone and FL with GM-CSF arms. The addition of FL to the vaccine regimen Was associated with the development of autoim­mune phenomenon in some patients. In general, the vaccine regimens including FL were well tolerated. One patient had grade 1 serologic abnormalities (antinuclear antibody, anti-SSA, anti-double-stranded DNA). The second patient, who had stage IV breast cancer, devel­oped grade 2 toxicity with serologic abnormalities and self-limiting Sicca syndrome characterized by dry eyes and dry mouth 3 months after the completion of the vaccine regimen. This patient did not develop any detectable immunity to HERZ peptides or protein after active immunization.

Summary

The cumulative data frum the limited number of com­pleted phase I clinical trials suggest that it is feasible to

use HERZ peptide-based vaccines in cancer patients to immunize against the HER2 protein. Future studies must now focus on maximizing immunity, developing strate­gies to generate potentially therapeutic antibody responses, and testing whether or not HER2-specific immunity affects tumor growth. Vaccination offers a therapeutic strategy that could potentially prevent the relapse of disease by establishing an effective memory response targeting HER2. Cancet vaccines t(lrgeting the HER2 oncogenic protein might be useful adjuvants to standard therapy and might aid in the prevention of relapse in patients whose tumors overexpress the protein.

Acknowledgments

This work is supported by grants from the National Insti tutes' of Heal lh, Nat ional Cancer .Institute (RO1 CA75 163 and CA85374, and KZ4 CA 85218) as welt as funding from the Cancer Research Treatment Founda­tion. We thank Ms. Chalie Livingston for assistance in manuscript preparation.

References

1. Siamon DJ. Clark GM, Wong SG, et al. Human breast cancer: correlation of relapse and survival with amplification o( the HER-2/neu oncogene, Science 1987; 2.35:177-82.

2. DLsis ML, Pupa SM, Grolow JR, et al. High-titcr HER-Z/neu proteuHpedftc antibody can be detected in patients with early­stage breast cancer. J elm Oneal 1997, 15:3363-7.

3. Disis ML. Knutson KL, Schiffman K, et al. rr~-ex\sre.nt immu­nity to [he HER-2/neu oncogenic pmtein in patients with HER­Uneu overcxpressing breast and ovarian cancer_ Breast Cancer Res Treat 2000; 62:245-52.

4. Ward RL, Hawk",s NJ, Coomber D, N a1. Antibody immUnity to the HER-2/neu oncogcruc protein in patients with colorectal cancer. Hum lmmun<Jl1999; 60:510-5.

5, Nagorsen D, Keilhol~ U, Rivoltini L, et at. Natural T-cell response against MHC class 1 epitopes of epithelial cell adhesion molecule, HER-Z/neu, and carcinoembryonic antigen. in patients with colorectal cancer. Cancer Res 2000; 60:4850-4.

6. McNeel DG, Nguyen LD, Storer BE, et aL Anribody immunity to prostate cancer associated antigens om be detected in the serum ofpatients with prostate cancer. J UTQ! 2000; 164:1825-9,

7. Zaks IT, Rosenberg SA. [mmunhation with a peptide epitop~ (p369-377) from HER-2/neu leads to peptide-specific cytotoxic T lymphocytes that fail to recognize HER-2{neu+ rumors. Cancer Res 1998; 58:4902-8.

B. Knutson KL, Schiffman K, Disis ML Immunization of cancer pa_ tiemswith a HER-l/neu HLA-A1 peptide, p369-377, result:; in shon-lived peptide specific immunity. elin Cancer Res 2002. In press.

9. Knurson KL, Schiffman K. Disis ML. Immuniwtion with a HER­l/neu helper peptide vaccine generat'" HER-2/neu COS T-cell iromunity in cancer patients.] CUn InllCSfW01; 107:477-84.

10. Disis ML, Gooley T, Rinn K, et at. Generation ofT cell immu· ni(y to the HER-2/neu protein after active immunization widl HER-2/neu peptide-based vaccines. ) Clin OncollOO2.1n press.

11. Brossart: P, Wirths S, Stuhler G, et al. Induction of cytotoxic T­lymphocyte responses in vivo after vaccinations with peptide­pu!sed dendritic cells. B100d 2000, 96:3102-S.

12. Disis ML, Rinn K, Knutson K, -e.t al. Fltl ligand as a vaccine adjUvant in a:rocw!ion with HER-2jneu peptide based vaccines in patients with HER-2/neu oveteX{lT=ing cancers. Btood 2002. In pres>.

13, Amon R, Van Regenmortcl MH. Structural hasis of antigenic spec,ficity and design of new vaccines. FASEB} 1992; 6:3265-74.

14. Disi. ML, Gralow JR, Bemharo H, et al. Peptide-based, but not whole protein, vaccines elicit immunity to HER.-2/neu, onco­genic self-protein.} Immuno! 1996; 156:3151-8,

15. Cibotti R, Kanellopoulos JM, Cabaniols JP, et al. Tolcrnnce to a self-protein involves its immunodominant but does not involve its subclominam dete.rmirumts. Proe Narl Acad Sci USA 1992; 89~416-20.

16. Sercan EE, Lehmann PV, Ametani A, et al. Dominance and crypticity of T cell antigemc detenn(nants. Annu Rev lmmlrn<ll 1993; 11 :729,66.

17. Nanda NK, Sercarl EE. Induction of anti-self-Immunity to cure cancer. Cd! 1995; 82:1.3-7.

18. Fisk B, Blevins TL, Wharton JT, et al. Identification of an imnmnodomlnam peptide of HER-2/neu protooncogene recog­nized by ovarian tumor-specific cytotoxic T lymphocyte lines. ] Exp Me.:! 1995; 181:2.109-17.

19. Disis ML, Bernhard H, Shiota FM. et a1. Granulocyte-macro­phage colony-stimulating factor: an effecrive adiuvant for protein and peptide-based vaccines. Blood 1996; 88:202-10.

20. Lehmann. PV, Fomhubcr T, Miller A, et al Spreading of T-cell autoimmunity w cryptic determinant,; of an autoantigen. Noture 1992; 358~ 155-7.

21. Se rca nEE. Driver clones and determina 0 t spreadmg. ] Auwimmun 2000; 14:275-7.

12. Marnskovsk.y E, Daro E, Raux E, et 31. In vivo generation of buman dendntlc cell subset:s by Flu ligand. Bload 2000: 96:878-84.

23. Pulendran B, Smith )L, Jenkins M, ot al. Prevention of peripheral tolerance by a clendmie cell growth factot: Flu ligand 35 all adju· vant, ] £4' Moo 1998; 18802075-81.