CHwERJG

DTDeiTct

R2FPHaCDBCgF

P

linical and Immunologic Responses ofLA-A3� Breast Cancer Patients Vaccinatedith the HER2/neu-Derived Peptide Vaccine,75, in a Phase I/II Clinical Trial

itesh Patil, MD, Guy T Clifton, MD, Jarrod P Holmes, MD, Asna Amin, MD, Mark G Carmichael, MD,eremy D Gates, MD, Linda H Benavides, MD, Matthew T Hueman, MD, Sathibalan Ponniah, PhD,eorge E Peoples, MD, FACS

BACKGROUND: We have treated disease-free breast cancer patients with an HER2/neu-derived peptide, E75, as anadjuvant vaccine. E75 was originally described as HLA-A2�restricted and has been previously testedin this population. Based on computer modeling, E75 is predicted to bind to HLA-A3, and preclin-ical data support this. We conducted a clinical trial of E75 in HLA-A3�, A2� (A3) patients.

STUDY DESIGN: Disease-freebreast cancerpatientswereenrolledafter standard therapy inphase I/II trials.A3patientswereenrolled in parallel with A2 patients and vaccinated with E75 and granulocyte-macrophage colony-stimulating factor immunoadjuvant. A2�, A3� patients were followed as controls. Toxicities weregraded. Immunologic responses were assessed by delayed-type hypersensitivity reactions and E75-specificinterferon-� enzyme-linked immunosorbent spot assay. Clinical recurrences were documented.

RESULTS: Thirteen A3 patients completed the vaccine schedule. Clinicopathologic features were similarbetween A3, A2, and control patients, except for more HER2/neu-overexpressing tumors in theA2 group and more estrogen-receptor/progesterone-receptor�negative tumors in A2 and A3groups. Toxicity profiles and postvaccination delayed-type hypersensitivity were similar in A3and A2 patients. Enzyme-linked immunosorbent spot assay results varied, but A3 patients’median spots increased pre- to postvaccination (p � 0.2). Recurrences were lower in the A3group (7.7%) at 30-month median follow-up compared with published recurrence in A2-vaccinated (8.3%) and control groups (14.8%) at 26-month median follow-up.

CONCLUSIONS: HLA restriction limits potential use of peptide-based cancer vaccines. This trial demonstratesthat HLA-A3 patients respond similarly to E75 vaccination as HLA-A2 patients, suggesting thepotential use of the E75 vaccine in up to 76% of the population. (J Am Coll Surg 2010;210:

140–147. Published by Elsevier Inc. on behalf of the American College of Surgeons)

Tastptu(Cdttsioa

isclosure Information: Dr Peoples holds patent rights to the E75 vaccine.his study was supported by the United States Military Cancer Institute,epartment of Surgery, Uniformed Services University of the Health Sci-

nces, and the Department of Clinical Investigation, Walter Reed Army Med-cal Center, Bethesda, MD.his article represents the personal viewpoint of the authors and cannot be

onstrued as a statement of official Department of the Army, Department ofhe Navy, or Department of Defense policy.

eceived August 26, 2009; Revised October 20, 2009; Accepted October 21,009.rom the Joyce Murtha Breast Care Center, Windber Medical Center, Windber,A (Patil); Department of Surgery, Brooke Army Medical Center, Fort Samouston, TX (Clifton, Gates, Benavides, Peoples); Department of Hematology

nd Medical Oncology, Naval Medical Center, San Diego, CA (Holmes); and theancer Vaccine Development Program, United States Military Cancer Institute,epartment of Surgery, Uniformed Services University of the Health Sciences,ethesda, MD (Amin, Carmichael, Hueman, Ponniah, Peoples).orrespondence address: COL George E Peoples, MD, FACS, Department of Sur-

ery, General SurgeryService,BrookeArmyMedicalCenter,3851RogerBrookeDr,

lort Sam Houston, TX 78234. email: george.peoples@amedd.army.mil

140ublished by Elsevier Inc. on behalf of the American College of Surgeons

umor-associated antigens (TAA) are a promising area ofctive research for cancer immunotherapy. Cancer vaccine re-earch has centered on priming the body’s immune system sohat it will recognize and specifically attack cancer cells ex-ressing TAAs. There have been multiple approaches usedo prime the immune system to these antigens. We havesed single-peptide vaccines given with immunoadjuvantgranulocyte-macrophage colony-stimulating factor [GM-SF]) in the adjuvant setting to patients who are surgicallyisease free and after completion of standard chemo/radiationherapy. Advantages of this method include a lack of substan-ial toxicity in addition to simplicity; low cost of production;tability in storage; and ability to generate a strong, specificmmune response that is easily monitored. The disadvantagef this method is that single peptides are often HLA-restrictednd could have limited applicability.

HER2/neu, a proto-oncogene expressed in many epithe-

ial malignancies, is one of the most extensively investigated

ISSN 1072-7515/10/$36.00doi:10.1016/j.jamcollsurg.2009.10.022

TowcdHdbtb

mccMpHrMoAMtnAwbasap

rhgWbnp

raawg

puibcm(fsaasHpppvvcc

MPNRiIAproicsprbat(ma

VT

141Vol. 210, No. 2, February 2010 Patil et al Clinical and Immunologic Responses with the E75 Peptide Vaccine

AAs. HER2/neu overexpression is found in 20% to 25%f breast cancers and is associated with aggressive diseaseith a worse prognosis.1 Several immunogenic peptides

apable of inducing an in vitro immune response have beenescribed from the HER2/neu protein. E75 (KIFGSLAFL,ER2/neu: 369-377) is the most studied HER2/neu-

erived peptide in laboratory and clinical studies.2-5 E75inds to HLA class I molecules to stimulate CD8� cyto-oxic T lymphocytes (CTLs) and is classically described aseing HLA-A2�restricted.There is still much that needs elucidating in the role ofajor histocompatibility complex (MHC) molecules in

ancer immunosurveillance and antitumor responses, butlearly it plays an important role. Expression levels of

HC class I molecules are an independent predictor ofrognoses in breast, ovarian, and colorectal cancers.6-8

LA-A2 (A2) in particular appears to play an importantole in epithelial cancers. HLA-A2 is the most common

HC class I antigen in North America (present in 49.5%f whites, 33.7% of blacks, and 55.9% of Asians).9 HLA-2�restricted TAAs have been identified from CEA,AGE, and HER2/neu.10-14 In addition, studies from

umor lymphocytes in HLA-A2� breast, ovarian, mela-oma, and gastric cancers have demonstrated HLA-2�restricted tumor cell lysis that is reproducible onlyith HLA-A2� allogenic tumor cells and can be inhibitedy monoclonal antibodies against HLA class I or HLA-A2ntigens.15-18 Interestingly, HLA-A2 has not been demon-trated to affect the incidence or prognosis of breast cancer,lthough this specific HLA type has been implicated as arognostic factor for advanced ovarian cancer.19-21

Our group recently published a phase II trial that en-olled 187 surgically disease-free breast cancer patients atigh risk for recurrence in a vaccine trial with an immuno-enic peptide, E75, with GM-CSF as an immunoadjuvant.e have demonstrated the safety and efficacy of the E75-

ased vaccine in node-positive (NP) and high-risk node-egative (NN) breast cancer patients. Initial results of the

Abbreviations and Acronyms

CTL � cytotoxic T lymphocyteDTH � delayed-type hypersensitivityELISPOT � enzyme-linked immunosorbent spot assayGM-CSF � granulocyte-macrophage colony-stimulating

factorIFN � interferonNN � node-negativeNP � node-positivePBMC � peripheral blood mononuclear cellsTAA � tumor-associated antigen

hase I/II trial of E75 showed significantly decreased recur- m

ences (5.6% versus 14.2%; p � 0.04) with only grades 1nd 2 toxicities at median 20-month follow-up, but thispparent clinical effect was lost when immunity wanedithout boosters.22 The E75 vaccine is currently being tar-eted to enter phase III trials for drug development.

E75 was originally described as an HLA-A2�restrictedeptide.23 We initially vaccinated HLA-A2� patients andsed HLA-A2� patients as controls. Computer modelingnformation from 2 commonly used HLA-peptide–inding algorithms, BIMAS24 and SYFPEITHI,25 indi-ated that HLA-A3 (A3), which is present in approxi-ately 2% to 26% of the population in the United States

white 26%, black 17%, Asian 2%), has binding affinityor E75.9 In addition, Sotiropoulou and colleagues demon-trated that HLA-A3� patients with breast, lung, ovarian,nd colorectal cancer had CTLs that recognized E75.26 Inddition, our group conducted preclinical studies thathowed that E75-stimulated HLA-A3� CTL could lyseLA-A3� HER2/neu-expressing cancer cells (data not

ublished). We then expanded the trial to incorporate A3atients in a study parallel to that being carried out in A2atients. In this article, we specifically report on the E75-accinated A3 patients in terms of toxicity, ex vivo and inivo immunologic responses, and clinical outcomes andompare these results with the vaccinated A2 patients andontrols.

ETHODSatient characteristics and clinical protocolsP and NN trials were approved by the local Institutionaleview Boards and conducted at Walter Reed Army Med-

cal Center and Joyce Murtha Breast Care Center under annvestigational New Drug application (BB-IND #9187).ll patients had biopsy-proven breast cancer and com-leted a standard course of surgery, chemotherapy, andadiation therapy (as required) before enrollment. Patientsn hormonal therapy were continued on their specific reg-men. After proper counseling and consenting, breast can-er patients were HLA-typed. In the initial phases of thetudy, HLA-A2� patients were vaccinated and HLA-A2�

atients were observed as controls prospectively for clinicalecurrence. When it was later established that E75 alsoinds to HLA-A3, NP and NN HLA-A3� patients werelso enrolled in parallel to A2. Before vaccination, all pa-ients were skin-tested with a panel of recall antigensmumps, tetanus, and Candida) and were considered im-unocompetent if they reacted (� 5 mm) to 2 or more

ntigens.

accinehe E75 peptide was commercially produced using good

anufacturing practices by NeoMPS, Inc. Peptide purity

(tdsiowtA

VPH2iDso(DppaHdGwtsed

TPdhtIEftgm

PaB(cVb

wips

EFwctsomtAdoeEw�ifR

DIrsvsssop

CAdpbc

SRvtcoiD

142 Patil et al Clinical and Immunologic Responses with the E75 Peptide Vaccine J Am Coll Surg

� 95%) was verified by high-performance liquid chroma-ography and mass spectrometry. Amino acid content wasetermined by amino acid analysis. Sterility and generalafety testing was carried out by the manufacturer. Lyoph-lized peptide was reconstituted in sterile saline at 100, 500,r 1,000 ug doses of E75 in 0.5 mL. The peptide was mixedith 0.5 mL GM-CSF (Berlex) and the 1.0-mL inocula-

ion was split and given intradermally at 2 sites 5 cm apart.ll inoculations were given in the same extremity.

accination serieshase I of the E75 vaccine trial (NP trial) included onlyLA-A2 NP patients. The NP trial was designed as a

-stage safety trial, with escalating doses of peptide in thenitial stage and alterations of schedule in the latter stage.

etails of the vaccine series have been published previou-ly.22 Briefly, 3 to 6 patients were each assigned to receive 4r 6 monthly injections of 100, 500, or 1,000 �g E75100.6; 500.4; 500.6; 1,000.4; and 1,000.6, respectively).uring the phase II portion of the trial, groups were ex-

anded to determine and confirm optimal dosing in NPatients, and enrollment of HLA-A3 patients was initiatedt this time. The second phase of the trial also included NNLA-A2 patients. The NN trial was designed to further

elineate optimal biologic dosing by varying the dose ofM-CSF and altering the inoculation schedule. Patientsith non-HER2/neu-expressing tumors were allowed in

his trial to determine the feasibility of vaccinating a pre-umably antigen-naive host. Ten patients were assigned toach dose group to receive 3, 4, or 6 monthly injectionsuring 5 months.

oxicityatients were observed 1 hour postvaccination for imme-iate hypersensitivity and returned 48 to 72 hours later toave their injection sites measured and be questioned aboutoxicities. Toxicities were graded by the National Cancernstitute Common Terminology Criteria for Adversevents v3.0 and reported on a scale from 0 to 5. Progression

rom 1 dose group to the next occurred only if no substan-ial toxicity (grade 4 or grade 5) occurred in the lower-doseroup. Patient-specific results are reported based on maxi-al local and systemic toxicity occurring during the series.

eripheral blood mononuclear cell isolationnd cultureslood was drawn before each vaccination and at 1 month

postvaccine) and 6 months (long term) after vaccine seriesompletion. Fifty milliliters of blood was drawn into BDacutainer CPT tubes (BD Biosciences), and peripheral

lood mononuclear cells (PBMC) were isolated. PBMCs d

ere washed and resuspended in culture medium consist-ng of RPMI with 5% heat-inactivated fetal calf serum andenicillin, streptomycin, and L-glutamine and used as aource of lymphocytes, as described previously.22,27,28

nzyme-linked immunosorbent spot assayreshly isolated PBMCs in culture medium supplementedith interleukin-7 (20 ng/mL) were added at 5 � 105

ells/200 �L/well into flat-bottom wells of antihuman in-erferon (IFN)-� enzyme-linked immunosorbent spot as-ay (ELISPOT) plates (BD Pharmingen). Duplicate wellsf the PBMC were stimulated in the absence (with cultureedium only) or presence of E75 at 25 �g/mL by placing

he plate overnight in an incubator (37°C � 5% CO2).fter overnight incubation, the plate was processed andeveloped as suggested by the manufacturer. The numberf spots (representing IFN-��secreting cells) present inach well was scanned and enumerated using the CTLLISPOT analyzer and the Immunospot Professional Soft-are (CTL Analyzers LLC). E75 peptide-specific IFN-�secreting cells (IFN spots) were calculated by subtract-

ng the total number of spots in the culture-medium wellsrom the total number of spots in the E75-stimulated wells.esults are presented as number of spots per 106 PBMCs.

elayed-type hypersensitivityn both the groups, a delayed-type hypersensitivity (DTH)eaction was assessed with 100 �g E75 in 0.5 mL sterilealine (without GM-CSF) and 0.5 mL sterile saline as aolume control 1 month after completion of the vaccineeries, as described previously.22 DTH reaction was mea-ured in 2 dimensions at 48 to 72 hours by using theensitive ballpoint-pen method and reported as the orthog-nal mean and compared with control.29 In the NN trial, arevaccination DTH was performed as well.

linical recurrencesll patients were observed for clinical recurrence per stan-ard of care cancer screening, as dictated by the patient’srimary oncologist. A patient was considered recurrent ifiopsy-proven or treated for recurrence by the primary on-ology team.

tatistical analysisecurrence rates were compared between groups using sur-ival analysis by the Kaplan-Meier method and the propor-ion of subjects using log-rank analysis. The p values forlinicopathologic factors were calculated using Wilcoxonr chi-square tests as appropriate. The p values for compar-ng prevaccination and postvaccination dimer levels and

TH reactions were calculated using Wilcoxon and Stu-

ent’s t-test as appropriate.

RDAAtewpcawpc

TFt

vbsrls(4(2ow09Arpgsr

DTi0(3ni

IEsPdp

FA(itet

TC

AWTHNECXHH

E*

143Vol. 210, No. 2, February 2010 Patil et al Clinical and Immunologic Responses with the E75 Peptide Vaccine

ESULTSemographics3 patients were enrolled in our vaccine trial parallel to the2 patients after it was determined from preclinical studies

hat E75 can bind HLA-A3. Overall, 187 patients werenrolled in the study; 93 A2 patients and 13 A3 patientsere vaccinated and 81 A2� (later A2- and A3-negative)atients were followed prospectively as controls. Standardlinicopathologic features were similar between A3, A2,nd control patients (Table 1), except that control patientsere considerably less likely to be estrogen-receptor/rogesterone-receptor�negative, and A2 patients had aonsiderably greater proportion of HER2/neu overexpression.

oxicityigure 1 depicts the maximum toxicity experienced duringhe vaccine series. Toxicities did not limit the completion of

igure 1. Maximum local and systemic toxicity experienced by HLA-2� and HLA-A3� patients. Maximal local and systemic toxicitydefined by Common Terminology Criteria for Adverse Events, v3.0)n A2 and A3 patients is depicted during the entire primary vaccina-ion series. Results are expressed as a proportion of the groupsxperiencing a given grade of toxicity. All rates were similar between

able 1. Demographic and Prognostic Characteristics of Vaharacteristic Control (n � 81)

ge, y 56hite, % 81.5

umor size (T2-4), % 39.5istologic grade 3, % 39.5ode-positive, % 56.8R/PR-negative, % 17.3hemotherapy, % 72.8RT, % 81.3ormonal therapy, % 77.5ER2/neu overexpression, % 26.0

R/PR, estrogen-receptor/progesterone-receptor; XRT, radiation therapy.Statistically significant.

mhe A2 and A3 groups.

accination series in any patient. Most of the toxicities inoth the groups were limited to expected grade 1 local andystemic reactions. In A3 patients, only 1 grade 2 systemiceaction was noted in an NP patient, ie, chest and back painasting for 2 hours. All other systemic reactions were tran-ient, grade 1 constitutional symptoms, including fatiguen � 10; 77%), headache (n � 9; 69%), arthralgias (n � 6;6%), myalgias (n � 5; 38%), nausea (n � 4; 31%), chillsn � 3; 23%), bone pain (n � 3; 23%), and back pain (n �; 15%). There were no grade 4 or 5 toxicities in either A2-r A3-vaccinated patients. Overall, the toxicity profilesere similar between A2 and A3 patients, except for gradeand grade 1 systemic reactions. There were 14.7% (14 of5) grade 0 reactions in A2 patients versus 0% (0 of 12) in3 patients (p � 0.21) and 70.5% (67 of 95) grade 1

eactions in A2 patients versus 90.4% (11 of 12) in A3atients (p � 0.17). The similarly high rate of expectedrade 1 local reactions between the A2 and A3 patientsuggests that the vaccine produces comparable immuneeactions in these 2 groups.

TH responsehe postvaccination DTH immune response was increased

n both groups. In the A3 group, the DTH increased frommm prevaccination to 10.7 � 2.5 mm postvaccination

p � 0.01). In the A2 group, the DTH increased from.5 � 0.8 to 14.5 � 1.3 mm (p � 1 � 10�6). There wasot a significant difference between postvaccination DTH

n the A3 and A2 patients (p � 0.2) (Fig. 2).

mmunologic responsex vivo ELISPOT assay to detect functional immune re-

ponses to the vaccine was performed using freshly isolatedBMCs, as described in the Methods section, and resultsemonstrated an overall increase in IFN-��secreting E75eptide-specific T cells. Three patients demonstrated a

ted A2, A3, and Control PatientsA2 (n � 93) A3 (n � 13) p Value

60 60 0.5189.2 100.0 0.2431.2 23.1 0.5137.8 57.1 0.9648.4 61.5 0.5932.6 30.8 0.04*74.2 76.9 0.9971.0 84.6 0.3364.5 69.2 0.2332.6 18.2 0.01*

ccina

arked increase in number of spots (Fig. 3A), and the

rc3scitEac

CBsfmppgca2

Fpbsvivnscfad(

FafbshtofTm

144 Patil et al Clinical and Immunologic Responses with the E75 Peptide Vaccine J Am Coll Surg

emainder showed a modest increase (4 patients), nohange (2 patients), or decrease (2 patients) (Fig. 3B andC). One patient with a likely spurious decrease of 160pots/106 cells prevaccination to 0 spots/106 cells is ex-luded from Figure 3 for clarity, and another patient is notncluded in the ELISPOT analysis because her prevaccina-ion sample had clotted. Overall, median prevaccinationLISPOT response was 7 spots/106 cells (range 0 to 160),nd median postvaccination response was 13 spots/106

ells (range 0 to 560; p � 0.20).

linical recurrencesecause the A3 group was added late, median follow-up is

ubstantially shorter for this group. Current medianollow-up is 30 months in the study for the A3 group, 41onths for the A2 group, and 43 months for the control

atients. The recurrence rate in the A3 group was com-ared with previous analysis. The recurrence rate in the A3roup is 7.7% at a median of 30 months of follow-upompared with 8.3% for the vaccine group (A2 and A3)nd 14.8% for the control group (p � 0.52) at a median of

igure 2. Delayed type hypersensitivity (DTH) reaction before andfter vaccination in A2 and A3 patients. DTH reactions were per-ormed with 100 mg peptide or an equal volume saline controlefore vaccination and 1 month after completion of the vaccinationeries. DTH reactions were measured in 2 dimensions at 48 to 72ours by using the sensitive ballpoint-pen method and reported ashe orthogonal mean. All patients’ results are shown with bar dem-nstrating the median value. There was an increase in DTH sizerom prevaccination to postvaccination in both A2 and A3 groups.here was not a statistical difference in the postvaccination DTHeasurements between the A2 and A3 groups.

6 months (Fig. 4).30s

igure 3. E75-specific interferon-� (IFN-�)�secreting cells in A3atients. Pre- and postvaccination IFN-� enzyme-linked immunosor-ent spot assay (ELISPOT) was performed in the A3 patients. Re-ults are expressed as the number of spots/106 cells. (A) Threeaccinated patients demonstrated a marked increase in peptide-nduced spot formation postvaccination series. The remainder ofaccinated patients showed (B) a modest increase (4 patients), (C)o change (2 patients), or decrease (2 patients) in the number ofpots. One patient with a likely spurious decrease of 160 spots/106

ells prevaccination to 0 spots/106 cells is excluded from Figure 3or clarity, and 1 other patient is not included in the ELISPOTnalysis because her prevaccination sample clotted. Overall, me-ian prevaccination ELISPOT response was 7 spots/106 cellsrange 0 to 160) and median postvaccination response was 13

pots/106 cells (range 0 to 560; p � 0.20).

DIapatbs

pocthtaAccttpAp

pippbmuibn

pidtepttcRppbamgrn

EgetcmCpiewocip

mmeacaevewvira

Fcor2

145Vol. 210, No. 2, February 2010 Patil et al Clinical and Immunologic Responses with the E75 Peptide Vaccine

ISCUSSIONn this study, we have shown that the E75 vaccine is safend effective at raising HER2/neu immunity in HLA-A3atients. The toxicity profile, local reactions, and in vivond in vitro immunologic responses are similar to A2 pa-ients. Most importantly, it appears that the vaccine mighte having a clinical impact in A3 patients similar to thateen in A2 patients.

The 13 A3 patients presented here are part of the largerhase I/II trial designed to investigate safety, define theptimal dose, monitor immunologic response, and testlinical efficacy of the E75 peptide vaccine as adjuvantherapy for clinically disease-free breast cancer patients atigh risk of recurrence. As stated previously, E75 was ini-ially thought to be HLA-A2�restricted and A2 patientslone received the vaccine and all others served as controls.fter evidence suggested that the HLA-A3 allele is alsoapable of presenting E75, HLA-A3 patients were also vac-inated and followed concurrently along with HLA-A2 pa-ients. The A3 patients had similar findings as the A2 pa-ients, including lower recurrence rates than the controlatients. Results extend the potential use of E75 to either2 or A3 patients, thereby covering 58% to 76% of theopulation with this single peptide vaccine.9

Some have argued against the individual peptide ap-roach to cancer vaccines because of the HLA restrictionssue in favor of whole tumor cell, whole protein, longeptide, or tumor DNA vaccination approaches. These ap-roaches are considered preferable, because they allow theody’s immune system to process antigens and present im-unogenic epitopes that are compatible with the individ-

al’s HLA type. In practice, a much smaller amount ofmmunogenic antigen is presented to the immune system,ecause the majority of the vaccine material is composed of

igure 4. Recurrence rates in A2, A3, and control patients. Clini-ally detected recurrences in the A3 patients measured at a medianf 30 months was 7.7% compared with previously reported recur-ence rates in the vaccine arm (A2 and A3) and control patients at6 months of 8.3% and 14.8%, respectively (p � 0.52).

ormal cellular antigens or nonimmunogenic material es- a

ecially for the whole tumor cell strategy. Importantly, themmune system machinery might preferentially process theominant epitopes of proteins to which T cells are moreolerant and can miss potentially effective subdominantpitopes.31 All of these approaches require the antigen-resenting cells to cleave, process, and present epitopes tohe adaptive immune system, which creates additional, po-entially rate-limiting steps. In the case of DNA vaccines,ells also have to acquire and transcribe DNA and translateNA to produce the antigens. All of these steps introduceotential variability into the vaccines. Finally, there areractical issues, such as safety and FDA approval, that muste addressed for these vaccination approaches to be export-ble to the community. The expense of these approachesight be greater, particularly for whole tumor cell strate-

ies, with irradiation and gene manipulation needed toemove the malignant potential or increase the immunoge-icity of the inoculated cells.32

Results observed in the current study suggest that the75 peptide vaccine has the ability to stimulate immuno-enic responses in A3 patients. In vivo clinical immunitylicited to the vaccine was clearly evident by the substan-ially increased DTH responses that were observed in vac-inated patients. Lack of availability of HLA-A3 dimerolecules prevented us from enumerating E75-specificD8 T cells in these patients, as was done with HLA-A2atients in our previous trials.22,27,28 We were able to mon-

tor E75-specific functional immune responses by using anx vivo ELISPOT assay, where freshly isolated PBMCsere stimulated directly in ELISPOT plates for a shortvernight incubation period. Using this approach, an in-reased number of IFN-��secreting cells were detectablen the E75-stimulated PBMCs when comparing pre- andostvaccination blood samples.The isolated peptide approach remains the simplest andost concise method of vaccinating cancer patients. Itight be effective, because it allows an epitope to be deliv-

red at a sufficiently high dose to overcome immune toler-nce to self-proteins and tolerance mechanisms of the can-er itself. Peptides can bind directly to HLA molecules orre internalized and presented in the context of HLA mol-cules, with minimal to no additional processing in-olved.33,34 In addition, the peptide can create a broaderffect through intra- and interantigenic epitope spreading,hich has been observed with the E75 vaccine.35 From aaccine development standpoint, it is easier to monitor themmunologic and toxicity effects from a single peptide andelatively inexpensive and easy to produce, making thispproach more easily exported to the community.

Ultimately, as new immunogenic epitopes become avail-

ble, peptides with varying HLA restrictions can be deliv-

estimtnpmtcttTi

rtMaliMatpfct

pdMtbrspt

A

SA

A

D

C

R

1

1

1

1

1

1

1

146 Patil et al Clinical and Immunologic Responses with the E75 Peptide Vaccine J Am Coll Surg

red as a mixture to all patients or separately as HLA-pecific vaccines. Inoculating with a mixture of peptideshat cover all HLA types allows one vaccine to be morenclusive. This approach has been used with 12 peptide

ixtures in melanoma in the adjuvant setting, covering upo 80% of patients.36 This approach would preclude theeed to HLA-type patients and allow manufacturers toroduce one single product. Such a mixture would deliverultiple ineffective peptides to individuals at the same

ime. Alternatively, the HLA-type�specific peptide vac-ine approach would only deliver peptides compatible withhe individual’s immune system but would require HLA-yping and manufacturing individual peptides separately.he latter approach fits with the current trends toward

ndividualized medicine.To produce the most effective immunologic and clinical

esponse, peptide vaccines might need to contain peptideshat target both MHC class I and II molecules. Although

HC class I peptides have been demonstrated to produceCD8� CTL response, this response is relatively short-

ived and requires boosting to maintain cellular immun-ty.37 The CD4�helper T lymphocytes, which are targeted by

HC class II molecules, enhance CD8� response, producentibody responses, and maintain long-lasting immunityhrough memory T cells.38 Combining MHC class I and IIeptides can produce a greater, more lasting anticancer ef-ect.39 Fortunately, MHC class II peptides are more promis-uous and might not require HLA-typing before administra-ion and might have more generalized applicability.

Our group has initiated studies examining a series ofeptides as cancer vaccines in the adjuvant setting withisease-free breast cancer patients.40 We are testing bothHC class I and II peptides separately and in combina-

ion. Ultimately, our goal is to use peptides as buildinglocks to construct the most effective vaccine for efficientlyaising a long-lived, antigen-specific protective immune re-ponse in an HLA-specific population. If these proof-of-rinciple trials work, these principles can be expanded tohe remainder of patients with new HLA-specific peptides.

uthor Contributions

tudy conception and design: Ponniah, Peoplescquisition of data: Patil, Clifton, Holmes, Amin, Car-

michael, Gates, Benavides, Huemannalysis and interpretation of data: Holmes, Hueman, Pon-

niah, Peoplesrafting of manuscript: Patil, Clifton, Benavides, Ponniah,Peoples

ritical revision: Patil, Clifton, Ponniah, Peoples

EFERENCES

1. Pritchard KI, Shepherd LE, O’Malley FP, et al. National CancerInstitute of Canada Clinical Trials Group: HER2 and respon-siveness of breast cancer to adjuvant therapy. N Engl J Med2006;354:2103–2111.

2. Anderson BW, Peoples GE, Murray JL, et al. Peptide priming ofcytolytic activity to HER-2 epitope 369-377 in healthy individ-uals. Clin Cancer Res 2000;6:4192–4200.

3. Zaks TZ, Rosenberg SA. Immunization with a peptide epitope(p369-377) from HER-2/neu leads to peptide-specific cytotoxicT lymphocytes that fail to recognize HER-2/neu� tumors.Cancer Res 1998;58:4902–4908.

4. Knutson KL, Schiffman K, Cheever MA, Disis ML. Immuniza-tion of cancer patients with a HER-2/neu, HLA-A2 peptidep369-377, results in short-lived peptide-specific immunity. ClinCancer Res 2002;8:1014–1018.

5. Murray JL, Gillogly ME, Przepiorka D, et al. Toxicity, immu-nogenicity, and induction of E75-specific tumor-lytic CTLs byHER-2 peptide E75 (369-377) combined with granulocytemacrophage colony-stimulating factor in HLA-A2� patientswith metastatic breast and ovarian cancer. Clin Cancer Res2002;8:3407–3418.

6. Madjd Z, Spendlove I, Pinder SE, et al. Total loss of MHC classI is an independent indicator of good prognosis in breast cancer.Int J Cancer 2005;117:248–255.

7. Rolland P, Deen S, Scott I, et al. Human leukocyte antigen classI antigen expression is an independent prognosis factor in ovar-ian cancer. Clin Cancer Res 2007;13:3591–3596.

8. Watson NF, Ramage JM, Madjd Z, et al. Immunosurveillance isactive in colorectal cancer as downregulation but not completeloss of MHC class I expression correlates with a poor prognosis.Int J Cancer 2006;118:6–10.

9. Lee TD. Distribution of HLA antigens in North American cau-casians, North American blacks and orientals. In: Lee J, ed. TheHLA system. New York: Springer-Verlag; 1990:154.

0. Alves PM, Viatte S, Fagerberg T, et al. Immunogenicity of thecarcinoembryonic antigen derived peptide 694 in HLA-A2healthy donors and colorectal carcinoma patients. Cancer Im-munol Immunother 2007;56:1795–1805.

1. Ras E, van der Burg SH, Zegveld ST, et al. Identification ofpotential HLA-A *0201 restricted CTL epitopes derived fromthe epithelial cell adhesion molecule (Ep-CAM) and the carci-noembryonic antigen (CEA). Hum Immunol 1997;53:81–89.

2. Zhang HG, Chen HS, Peng JR, et al. Specific CD8(�)T cellresponses to HLA-A2 restricted MAGE-A3 p271-279 peptidein hepatocellular carcinoma patients without vaccination. Can-cer Immunol Immunother 2007;56:1945–1954.

3. Dong HL, Li ZS, Ye J, et al. Identification of HLA-A2-restrictedCTL epitope encoded by the MAGE-n gene of human hepato-cellular carcinoma. Cancer Biol Ther 2004;3:891–898.

4. Lustgarten J, Theobald M, Labadie C, et al. Identification ofHer-2/Neu CTL epitopes using double transgenic mice express-ing HLA-A2.1 and human CD.8. Hum Immunol 1997;52:109–118.

5. Linehan DC, Goedegebuure PS, Peoples GE, et al. Tumor-specific and HLA-A2-restricted cytolysis by tumor-associatedlymphocytes in human metastatic breast cancer. J Immunol1995;155:4486–4491.

6. Peoples GE, Goedegebuure PS, Andrews JV, et al. HLA-A2

presents shared tumor-associated antigens derived from en-

1

1

1

2

2

2

2

2

2

2

2

2

2

3

3

3

3

3

3

3

3

3

3

4

147Vol. 210, No. 2, February 2010 Patil et al Clinical and Immunologic Responses with the E75 Peptide Vaccine

dogenous proteins in ovarian cancer. J Immunol 1993;151:5481–5491.

7. Darrow TL, Slingluff CL, Seigler HF. The role of HLA class Iantigens in recognition of melanoma cells by tumor-specificcytotoxic T lymphocytes: evidence for shared tumor antigens.J Immunol 1989;142:3329–3335.

8. Kono K, Rongcun Y, Charo J, et al. Identification of HER2/neu-derived peptide epitopes recognized by gastric cancer-specific cytotoxic T lymphocytes. Int J Cancer 1998;78:202–208.

9. Gourley C, Thornton C, Massie C, et al. Is there a relationshipbetween HLA type and prognostic factors in breast cancer? An-ticancer Res 2003;23:633–638.

0. Iaffaioli RV, Maio M, Ruggiero G, et al. HLA and prognosticfactors in primary breast cancer. Int J Cancer 1985;35:581–585.

1. Gamzatova Z, Villabona L, Dahlgren L, et al. Human leucocyteantigen (HLA) A2 as a negative clinical prognostic factor inpatients with advanced ovarian cancer. Gynecol Oncol 2006;103:145–150.

2. Peoples GE, Holmes JP, Hueman MT, et al. Combined clinicaltrial results of a HER-2/neu (�75) vaccine for the prevention ofrecurrence in high risk breast cancer patients. United States Mil-itary Cancer Institute Clinical Trials Group Study. I-01 andI-02. Clin Cancer Res 2008;14:797–803.

3. Fisk B, BlevinsTL, Wharton JT, Ioannides CG. Identification ofan immunodominant peptide of HER-2/neu protooncogenerecognized by ovarian tumor-specific cytotoxic T lymphocytelines. J Exp Med 1995;181:2109–2117.

4. Parker KC, Bednarek MA, Coligan JE. Scheme for ranking po-tential HLA-A2 binding peptides based on independent bind-ing of individual peptide side-chains. J Immunol 1994;152:163–175.

5. Rammensee HG, Bachmann J, Stevanovic S. MHC ligands andpeptide motifs (Molecular Biology Intelligence Unit). NewYork: Springer; 1997.

6. Sotiropoulou PA, Perez SA, Iliopoulou EG, et al. CytotoxicT-cell precursor frequencies to HER-2 (369-377) in patientswith HER-2/neu-positive epithelial tumors. Br J Cancer 2003;89:1055–1061.

7. Hueman MT, Dehqanzada ZA, Novak TE, et al. Phase I clinicaltrial of a HER-2/neu peptide (E75) vaccine for the prevention ofprostate-specific antigen recurrence in high-risk prostate cancerpatients. Clin Cancer Res 2005;11:7470–7479.

8. Woll MM, Fisher CM, Ryan GB, et al. Direct measurement of

peptide-specific CD8� T cells using HLA-A2:Ig dimer for

monitoring the in vivo immune response to a HER2/neuvaccine in breast and prostate cancer patients. J Clin Immunol2004;24:449–461.

9. Jordan TJ, Sunderam G, Thomas L, Reichman LB. Tuberculinreaction size measurement by the pen method compared to tra-ditional palpation. Chest 1987;92:234–236.

0. Mittendorf EA, Holmes JP, Ponniah S, Peoples GE. The E75HER2/neu peptide vaccine. Cancer Immunol Immunother2008;57:1511–1521.

1. Cibotti R, Kanellopoulos JM, Cabaniols JP, et al. Tolerance to aself-protein involves its immunodominant but does not involveits subdominant determinants. Proc Natl Acad Sci USA 1992;89:416–420.

2. Dols A, Meijer SL, Smith JW 2nd, et al. Allogeneic breast cancercell vaccines. Clin Breast Cancer 2003;3[Suppl 4]:S173�S180.

3. Sherman LA, Burke TA, Biggs JA. Extracellular processing ofpeptide antigens that bind class I major histocompatibility mol-ecules. J Exp Med 1992;175:1221–1226.

4. Vitiello A, Potter TA, Sherman LA. The role of beta2-microglobulin in peptide binding by class I molecules. Science1990;250:1423–1426.

5. Mittendorf EA, Gurney JM, Storrer CE, et al. Vaccination witha HER2/neu peptide induces intra- and inter-antigenic epitopespreading in patients with early stage breast cancer. Surgery2006;139:407–418.

6. Slingluff CL Jr, Petroni GR, Chianese-Bullock KA, et al. Immu-nologic and clinical outcomes of a randomized phase II trial oftwo multipeptide vaccines for melanoma in the adjuvant setting.Clin Cancer Res 2007;13:6386–6395.

7. Holmes JP, Amin A, Storrer CE, et al. Clinical and immunologiceffects of a HER2/neu (E75) peptide vaccine booster in previ-ously vaccinated breast cancer patients [abstract]. J Clin OncolASCO Annu Meet Proc 2007;25:3014.

8. Klarnet JP, Matis LA, Kern DE, et al. Antigen-driven T cellclones can proliferate in vivo, eradicate disseminated leukemia,and provide specific immunologic memory. J Immunol 1987;138:4012–4017.

9. Disis ML, Grabstein KH, Sleath PR, Cheever MA. Generationof immunity to the HER-2/neu oncogenic protein in patientswith breast and ovarian cancer using a peptide-based vaccine.Clin Cancer Res 1999;5:1289–1297.

0. Holmes JP, Benavides LC, Gates JD, et al. Results of the firstphase I clinical trial of the novel Ii-key hybrid preventive HER2/

neu peptide (AE37) vaccine. J Clin Oncol 2008;26:3426–3433.