hsppc-96: a personalised cancer vaccine

Caudill & LiHSPPC-96: a personalised cancer vaccine

Oncologic, Endocrine & Metabolic

HSPPC-96: a personalised cancer vaccine

Marissa M Caudill & Zihai Li

Center for Immunotherapy of Cancer and Infectious Diseases, University ofConnecticut School of Medicine, USA

HSPPC-96 is a protein peptide complex consisting of a 96 kDa heat shockprotein (Hsp), gp96, and an array of gp96-associated cellular peptides.Immunisation with HSPPC-96 induces T-cell specific immunity againstthese peptides; gp96 is not immunogenic per se. The non-covalent bindingof gp96 to peptides is neither selective nor restricted to cell types. Thus, thecollection of gp96-associated peptides represents the antigenic peptidepool of a cell; this is the basis of the peptide-specific immunogenicityelicited by HSPPC-96. Since the repertoire of antigenic peptides in the cell isdifferent between normal and tumour cells and is also distinct amongtumours of the same histology due to the randomness of tumour-specificmutations, a purified sample of HSPPC-96 can only effectively immuniseagainst the tumour from which it originates. The immunisation effect is,therefore, specific to both the individual patient and the tumour type. Thistrue custom-based vaccine has been shown to induce protective T-cellimmunity against a variety of tumours in animal models. The majormechanism of action is thought to be the ability of HSPPC-96 to primepeptide-specific MHC class I-restricted CD8+ T-cells by interacting withantigen-presenting cells in a receptor-dependent manner. HSPPC-96 is nowbeing widely tested against human malignancies for which effectivetherapies do not exist. The major challenges of the human studies areimmunological monitoring and the determination of optimal dosing,scheduling and route of administration. HSPPC-96 has been well toleratedat all dose levels. Clinical applications are being addressed by numerousstudies including an international, multi-centre, double-blind Phase III trialfor the treatment of stage III renal cell carcinoma in the adjuvant setting.

Keywords: cancer vaccine, gp96, heat shock protein, HSPPC-96

Exp. Opin. Biol. Ther. (2001) 1(3):539-547

1. Introduction

Despite the tremendous advancements in cancer biology research, themain weapons against this devastating disease remain surgery, radiationtherapy and the rather non-selective cytotoxic drug therapy known custom-arily as chemotherapy. With a few exceptions, such as testicular cancer,childhood leukaemia and Hodgkin’s disease, the outcomes of most humanmalignancies are currently no better than they were three decades ago [1].

The renewed interest in the concept of immunotherapy against cancer isdue to the following three principles [2]:

• critics have not yet been able to disprove the concept that tumours arethe consequence of failed immunosurveillance by the immune system

5392001 © Ashley Publications Ltd. 1471-2598

Drug Evaluation

1. Introduction

2. Mechanism

3. General strategy forclinical applications

4. Dosage and schedule

5. Safety and tolerability

6. Clinical trials

7. Expert opinion

Bibliography

http://www.ashley-pub.com

Expert Opinion on Biological Therapy

www.ashley-pub.com

• tumours are indeed recognisable by the immunesystem because they express either tumour-specific antigens (such as viral antigens andoncogene products), or antigens that are preferen-tially expressed by the tumours (such as earlyembryonic antigens)

• the adaptive immune system can be manipulatedor optimised by vaccination, passive immunisationor modulation of the immunological climate bysystemic therapy with cytokines or other immuno-modulating agents

HSPPC-96 was first isolated in a murine fibrosarcomacell line called Meth A. In the process of identifyingtumour specific antigens capable of immunisingBalb/c mice to mediate rejection of Meth A [3], it wasshown that Meth A-derived gp96 but not gp96 fromanother tumour or from normal tissues, can immunisemice against challenge with live Meth A tumour cells.Since gp96 is expressed abundantly in both tumourcells and normal tissues, it was thought that tumourspecific mutations of gp96 were the basis of theobserved tumour-specific immunogenicity [4]. It tooknearly 10 years to realise that gp96 exists as a protein-peptide complex and the immunogenicity of gp96,like that elicited by other Hsps, is in fact due to associ-ated peptides [5,6]. No structural mutation of gp96itself has ever been identified. In addition, it is nowknown that gp96 is a bona fide protein chaperone andis a member of the Hsp family, which is involved inthe folding and unfolding of cellular proteins [5].

This paper aims to examine the principle of immuno-therapeutic strategy with HSPPC-96. It will highlightthe immunological properties of this molecule andupdate the experience so far in the clinical develop-ment against human malignancies.

2. Mechanism

2.1 HSPPC-96 carries antigenic peptides

gp96 is the most abundant protein in the lumen of theendoplasmic reticulum (ER) [7]. The role of gp96 infacilitating protein folding was suggested initially bythe finding that the introduction of misfolded proteinin the ER increases the expression of gp96. Like otherHsps, gp96 can interact directly with numeroussubstrates such as collagen, heparin, immunoglobulinheavy chain and glycoprotein of herpes simplex virus.The broad spectrum of substrates is consistent withthe notion that Hsps interact with their substrates

loosely through hydrophobic associations [8]. Thebinding of peptides to gp96 was confirmed by thefollowing studies. The first showed that disruption ofHSPPC-96 with a mild acid led to the liberation ofpeptides, which could be identified by structuralanalysis as well as immunological assays usingpeptide-specific cytotoxic T-lymphocytes (CTL)activity as a read-out [5,9,10]. Additionally, gp96isolated from cells containing known antigens (viralantigens, bacterial antigens or other model antigens)were able to immunise mice against these antigens;this was strong evidence that antigen-specificpeptides were present in gp96 preparations [9-12].Finally, it was found serendipitously that heattreatment of gp96 allowed the protein to bind withexogenous peptides, which could be measured by gelelectrophoresis or other biophysical means [13]. Bythis method, along with cross-linking, proteasemapping and microsequencing, a minimal peptide-binding site of gp96 was recently defined [14].

Despite these advances, several important questionsremain to be answered. The stoichiometry of gp96 topeptide binding is still unclear. Also, although gp96does bind to ATP and has intrinsic ATPase activity,ATP has not been shown to affect peptide binding togp96 in vitro (Z Li, unpublished). All efforts so far tounload gp96 in another physiological way have notmet with much success. Therefore, it is still unclearwhat prompts peptide dissociation in vivo. Related tothis, it is also unknown how the peptides associatedwith gp96 are eventually transferred to MHC class Imolecule in the cell [15].

2.2 HSPPC-96 interacts with antigen-presentingcells

HSPPC-96 immunisation has several unique features.It operates independently of adjuvant in mousemodels [3]. Moreover, the amount of peptide in thecomplex required to elicit T-cell immunity is severalorders of magnitude less than what would be requiredif free peptides were used alone [16]. The immunisa-tion effect of HSPPC-96 is also exquisitely dependenton phagocytic cells [17]. These features togethersuggest that HSPPC-96 might bind to a receptor-likemolecule on the surface of professional antigen-presenting cells (APC) which then presentgp96-bound peptides to induce optimal T-cellpriming [18]. Indeed, it was found that soluble gp96can stimulate dendritic cells, which are one of themost powerful APC types, to secretepro-inflammatory cytokines and to upregulate the

© Ashley Publications Ltd. All rights reserved. Exp. Opin. Biol. Ther. (2001) 1(3)

540 HSPPC-96: a personalised cancer vaccine

expression of co-stimulatory molecules such as CD86and MHC class II [19,20]. gp96 was shown to bindsome receptor-like molecules on the surface of APCsin a saturatable and competitive manner [21]. By usingaffinity purification with a gp96-conjugated column,Binder et al. were able to isolate a cell surface ligand(receptor) for gp96 from a macrophage cell line, RAW264.7, to homogeneity [22]. A polyclonal antibodyraised against this protein can block the representa-tion of gp96 chaperoned peptide to MHC class I,providing strong evidence that this could well be themain method of gp96-mediated peptide delivery toAPCs. Microsequencing of this molecule by massspectrometry confirmed it to be CD91, a proteinknown as α2-macroglobulin receptor or low densitylipoprotein-related protein (LRP). Further evidence tosupport CD91 as a receptor for gp96 is thatα2-macroglobulin, a previously known CD91 ligand,inhibi ts macrophages from re-present inggp96-chaperoned antigenic peptides [22]. However,CD91 is a complex molecule and its mechanism ofaction may not be so straightforward. There isevidence that CD91 may not be the only surfacemolecule that interacts with gp96. It has also beenfound that gp96 binds less efficiently to APC frommice that have homologous deletion of CD36, an 88kDa scavenger receptor, suggesting that CD36 may beanother receptor for gp96 [23].

2.3 HSPPC-96 primes CD8+ T-cells

HSPPC-96 immunisation can prime CD8+ T-cellsagainst many intracellular antigens, such as tumour,viral, minor histocompatibility and bacterial antigens.The depletion of various subsets of T-cells before andduring the immunisation period with Meth A-derivedHSPPC-96, led to the conclusion that the priming ofMeth A tumour-specific CD8+ T-cells by HSPPC-96 isindependent of help from CD4+ T-cells [17]. Interest-ingly, HSP70 fused with a model peptide can alsoprime CD8+ T-lymphocytes in a CD4+ T-cellindependent manner [24,25]. It is important to make adistinction between direct and cross priming of CD8+

T-cells. Direct priming strictly means direct activationof naïve T-cells by antigen expressing target cells,which mandates the presence of antigen in thecontext of the same MHC molecules in both the T-cellsand the target cells, in addition to the appropriateco-stimulatory molecules. Cross priming wasoriginally described as the priming of CD8+ T-cells bytarget cells that do not share the same MHC alleles asthe antigen specific T-cells [26]. In this case, the

antigens from the target cells have to be releasedexogenously and then cross-presented to the T-cellsby APCs which possess the proper MHC molecules.Recent data have confirmed that cross priming is thepredominant, if not exclusive, mode of T-cell activa-tion. Antigens from non-haematopoietic cells can bereleased and then taken up and presented by theprofessional APCs to T-cells [27]. The mechanism ofcross priming in the case of gp96 is not entirely clear.One possibility is the release of HSPPC-96, either fromcell death or stress [28]. The extracellular HSPPC-96may then interact with APCs through a specific gp96receptor such as CD91 or CD36. Two events thenensue. First, the peptides of HSPPC-96 are unloadedwithin the cell and eventually presented on MHC classI molecules. Second, APCs are activated to the‘priming mode’ due to increased cell surface expres-sion of co-stimulatory molecules and the secretion ofpro-inflammatory cytokines.

2.4 HSPPC-96 primes CD4+ T-cells

Can HSPPC96 present the peptides to MHC class IIpathway and thereby activate CD4+ T-cells?Sequencing by Edman degradation of the pooledpeptides from gp96 showed that the average length ofgp96-associated peptides is around 12 amino acids (ZLi, unpublished), although longer peptides of over 20amino acids can be successfully charged to gp96 invitro. Therefore, it is conceivable that MHC class II,which has the capacity to bind to longer peptides,may present some of the peptides associated withgp96. Thus far, however, researchers have not identi-fied any MHC class II-associated peptide epitopesfrom HSPPC-96. The only evidence that HSPPC-96may be able to prime CD4+ T-cells comes fromfunctional studies. Meth A-derived HSPPC-96generates CD4+ T-cells in immunised mice. Inaddition, APCs pulsed with Meth A HSPPC-96 canstimulate a Meth A-specific CD4+ T-cell clone toproduce cytokines (Matsutake T and Srivastava PK,personal communications). At high doses, HSPPC-96immunisation activates a subset of CD4+ T-cells thathave a suppressive phenotype [29]. It is unclear,however, whether or not these T-cells are antigen-specific.

2.5 Non-immunological function of HSPPC-96

Surprisingly, it has been shown that gp96 also binds toplatelets [30]. Although the binding does not seem tolead to any functional consequences in platelets, suchas adhesion or activation, it is prudent to keep in mind


Caudill & Li 541

that many possible functions of HSPPC-96 have notbeen completely characterised.

3. General strategy for clinical applications

Unlike traditional cancer drugs, HSPPC-96 vaccine isindividually based and truly tailored towards anindividual tumour from an individual patient [31]. Asearly as the 1940s, it was appreciated that effectivetumour immunity is exquisitely specific, i.e., usingtwo histologically identical tumours in mice, A and B,tumour A could only immunise against itself, notagainst tumour B and vice versa [28]. Thisdemonstrated clearly that tumours are antigenicallydistinct from one another. We can now speculate thatthis observed phenomenon is most likely due to thedifference in peptide pools as a result of randommutations in the transformed cell. Depending on thedifferent types of tumour, the grade, differentiationstage or genetic background, HSPPC-96 should alsobe expected to be individually unique, due to thedifferent composition of peptides associated withgp96. Therefore, in order to generate effectiveimmunity against tumour A, HSPPC-96 has to bedelivering the peptides specific to tumour A, nottumour B.

The general strategy is thus given the acronym‘OTHER’, meaning, stepwise, (a) Obtain tumour (bysurgery or other means) from the patient; (b)HSPPC-96 extraction in the laboratory; (c) Return (orreinjection) of HSPPC-96 to the patient. SinceHSPPC-96 purification is now standardised andtumours are often routinely excised for diagnostic ortherapeutic purposes, this truly custom based andunprecedented approach has become feasible. Morethan 200 patients have now been tested by thisOTHER approach worldwide.

4. Dosage and schedule

Drug development for use in human cancer treatmentmust progress through different phases. The primarygoals of each phase are:

• Phase I: to define maximum tolerated dose (MTD)

• Phase II: to monitor efficacy

• Phase III: to define long-term risks and benefits

HSPPC-96 is an autologous protein-peptide complex.It is not directly cytotoxic, so MTD may not be

definable in this case. In addition, high dose is clearlynot optimal for HSPPC-96, as shown in animal studies[29]. Using HSPPC-96 purified from a Meth A fibrosar-coma, it was found that the optimal dose to immunisemice against Meth A challenge varied depending onthe route of administration. The optimal doses forintradermal, sc. and ip. route were 1 µg, 10 µg and 50µg respectively [29]. As mentioned previously, doseshigher than the optimal dose can even cause antigen-specific immunosuppression in a manner that is stillunclear, although CD4+ T-cells seem responsible forthe suppression [29].

To determine the optimal human dose of HSPPC-96, aPhase I study was performed using patients with stageIV renal cell carcinoma. A total of 42 patients wereinitially enrolled in this study, of whom 38 hadenough HSPPC-96 purified from resected primarytumours to be included in the study. The study lookedat three different doses, 2.5, 25 and 100 µg of autolo-gous tumour-derived HSPPC-96 given id. weekly for atotal of four injections. There were no significanttoxicities observed, even at the 100 µg dose level.From this preliminary trial, the dosage of 25 µg perinjection seemed to be optimal. There were signifi-cantly more treatment failures in both the 2.5 µg and100 µg groups, underscoring the fact that HSPPC-96does have a narrow therapeutic window [32]. A similarstudy is being done for gastric cancer patients whohave had gastric surgery with curative intent. As ofMay 2000, 5/6 patients progressed after immunisationwith 2.5 µg id. weekly, whereas 3/4 patients remainedtumour free at dose level of 25 and 100 µg perinjection [33].

The schedule of HSPPC-96 administration is also notyet optimally defined. In the preclinical animalmodels, two weekly injections of tumour-derivedHSPPC-96 led to protection against subsequenttumour challenge. In contrast, tumour-bearinganimals have to be vaccinated with HSPPC-96 moreoften in order to slow down the tumour growth. Thereare at least two reasons to explain this difference [34].First, T-cells from tumour-bearing animals may betolerant to the tumour antigens since tumour cells donot express co-stimulatory molecules. Second, toogreat a tumour burden may out-weigh the antitumourresponses. Therefore, it is conceivable that multipleinjections of HSPPC-96 will have to be given, even inthe adjuvant setting, since microscopic metastases arecommon in cancer patients even when the primarydisease is no longer detectable.



5. Safety and tolerability

Theoretical risks of HSPPC-96 injections are:

• endotoxin-like reaction due to the secretion ofpro-inflammatory cytokines such as TNF-α as aresult of the activation of APCs [19,20]

• autoimmune phenomena as a consequence ofT-cell activation against the self-peptides inHSPPC-96

Studies have shown so far that these potential sideeffects do not occur. HSPPC-96 has been shown to bevery safe and well-tolerated. In the first pilot trial [35],16 patients with various advanced malignancies(seven gastrointestinal tract, one pancreatic, twohepatocellular, three thyroid, one breast, onemesothelioma and one unknown primary), which hadbecome refractory to established therapies, weretreated with autologous tumour-derived HSPPC-96.Patients were injected sc. with 25 µg autologousHSPPC-96 once a week for four weeks. After eachvaccination, patients were carefully monitoredclinically and serologically. No unacceptable toxicitiesor autoimmune phenomena were observed. Threepatients reported mild hot flushes between 30 - 48minutes after immunisation, without the need formedical intervention. Of these, two experienced hotflushes after each immunisation, while one experi-enced them again only after the second immunisation.In four patients, pain or fever occurred during theimmunisation period, however their onset wasunrelated to the timing of immunisation and theevents were attributed to tumour progression. Noneof these patients mounted significant titre of thefollowing auto-antibodies: antinucleoproteinantibody, antibody to dsDNA or ssDNA, antibodies tomitochondria, thyroglobul in, microsomes,perinuclear or cytoplasmic antigens after HSPPC-96immunisation. One patient died 4 days after thesecond immunisation, which was determined to bedue to the progression of his disease. Thus, at thispoint there is no evidence of acute HSPPC-96 inducedside effects due to self-destructive immune reactions.

Similarly, no significant adverse effects have beenobserved in subsequent trials on renal cell carcinoma,gastric cancer, colorectal cancer, melanoma, pancre-atic cancer or non-Hodgkin’s lymphoma.

6. Clinical trials

There are a total of 11 clinical trials on HSPPC-96 todate. These trials are classified into three categories.First is a pilot trial that has been completed andreported, which involved 16 patients with variouscancers treated with 25 µg HSPPC-96 [35]. The secondcategory includes five trials that are either closed(pancreatic cancer), or enrolment completed (dose -defining trial on stage IV renal cancer, HSPPC-96 plusIL-2 for renal cancer, dose-defining studies forcolorectal cancer and melanoma). The final categoryincludes five active trials that are still enrollingpatients. They are:

• one Phase III adjuvant therapy study for renal cellcarcinoma

• several Phase II s tudies for low-gradenon-Hodgkins’s lymphoma, gastric cancer,sarcoma

• a recently re-launched pancreatic cancer trial

It is important to point out that the primary objectivesof all of the trials except the Phase III renal carcinomatrials are feasibility and toxicity. Efficacy is not the endpoint for these studies. In addition, the follow-up forall studies is still not long enough to be conclusive.Therefore, it is premature and even prohibitive tocomment on the efficacy of HSPPC-96 at this earlystage of clinical development.

To comply with the ethical and scientific standard,only published results will be summarised.

6.1 The HSPPC-96 pilot study

In the pilot study, all patients had refractory metastaticdisease and had exhausted other therapeutic options[35]. Immunisation was performed weekly using 25 µgHSPPC-96 per dose. Of interest, six out of twelvepatients had increased MHC class I-restrictedIFN-γ-producing CD8+ T-cells specific against autolo-gous tumours or tumour membranes as measured byan IFN-γ ELISPOT assay. Moreover, eight out ofthirteen patients had expanded NK cell population byflow cytometric analysis of CD16+ and CD57+ cells.Clinically, four patients experienced stabilisation oftheir disease for 3 - 7 months although the details andsite of extent of diseases are not available. One patientwith advanced hepatocellular carcinoma was found tohave extensive tumour necrosis of over 50% of theprimary tumour after the third injection withHSPPC-96, accompanied by a strong and specific


Caudill & Li 543

anti-HCC CD8+ cell response. With exception to thispatient, no significant tumour lysis was observed inother patients. These data suggest that HSPPC-96 isable to stimulate T-cell immunity even in patients withadvanced stage malignancies.

6.2 Renal cell carcinoma

Renal cell cancer is a good model for immunologicalinterventions involving the use of tumours becausethe primary tumour burden is often large and theinitial diagnosis and treatment are radical nephrec-tomy, allowing for easy purification of HSPPC-96.Moreover, although there is no satisfactory treatmentfor this disease, spontaneous remission has beenobserved and a subset of patients can be cured withIL-2 and/or interferon.

The first Phase I renal cancer trial involved a total of 42stage IV patients treated in three dosage groups (2.5,25 and 100 µg per injection) [32]. A total of 38 patientswere evaluable. Clinical activity has beendemonstrated in 24% of patients for over 14 months,which compares favourably to the other availabletherapies for this population of patients. Full details inthe Phase I trial have not yet been published. A total of70 patients have been enrolled in a Phase II trial [36].All patients received four weekly 25 µg id. injectionsof HSPPC-96, followed by two additional 25 µg dosesat two week intervals. Responding or stable patientswere continued with four additional doses of 25 µg attwo week intervals; those patients who exhibiteddisease progression continued with vaccine every twoweeks, plus a 5 day per week course of 11 millionunits of IL-2 delivered subcutanenously for fourconsecutive weeks. As reported in October 2000 [37],25 patients have completed the therapy, of whom 8patients (32%) have continued on HSPPC-96 therapyalone. Of these eight, one patient is in completeremission, one has had a partial response and six havestable disease. Other patients have progressivedisease requiring addition of IL-2. Since this study isnot yet complete, it remains unclear whethercombination of IL-2 with HSPPC-96 is effective for thispopulation.

A multi-institute worldwide randomised Phase IIIstudy was launched in the summer of 2000 to studythe effect of autologous tumour-derived HSPPC-96vaccination in stage III renal cell cancer patients whohave undergone curative surgery. Relapse-freesurvival and overall survival will be the sole objectivesof this study.

6.3 Pancreatic cancer

HSPPC-96 was tested in a Phase I clinical trial forpatients with pancreatic cancer [38]. Patients who hadprimary, resectable pancreatic cancer were enrolledin the trial. After surgery to remove their primarytumour, patients were treated with 5 µg of vaccineintradermally. It was not possible to test higher doselevels due to the small size of tumours in thesepatients. The vaccine was prepared successfully from5/15 pancreatic cancer samples. It was not possible toprepare HSPPC-96 from the remaining specimens dueto the presence of proteolytic enzymes in the pancre-atic tissue that break down proteins, including Hsps.Tumour-specific T-cell mediated immune responseswere demonstrated in two of the five patients. Thesetwo patients were treated in 1997 and are still aliveand disease free in November 2000 (Lewis J, personalcommunication). A third patient was known to be freeof disease at 12 months and was subsequently lost tofollow-up. The fourth patient is alive with recurrentdisease at 10 months and the fifth patient died sevenmonths after starting HSPPC-96 treatment. Due tothese encouraging clinical results, the pancreatic trialhas now been expanded to treat more patients.

6.4 Gastric cancer

Patients with a diagnosis of gastric carcinomaunderwent gastric surgery with curative intent [33]. Atleast 1.5 grams of resected tumour was procured fromeach patient and HSPPC-96 was purified. Beginningfour to eight weeks after surgery, patients werevaccinated intradermally four to nine times in eitherweekly or two-week intervals with either 2.5, 25 or100 µg of HSPPC-96. Vaccine preparation wassuccessful in 12 out of 15 patients, showing the greaterfeasibility of vaccine preparation from resected gastriccarcinoma than pancreatic cancer. As of May 2000, sixpatients received 4 x 2.5 µg, two patients 4 x 25 µg,one patient 8 x 25 µg and one patient 9 x 100 µgHSPPC-96. All five stage IIIb and IV patients and onestage IIIa patient developed tumour progression aftera median of 6.5 months after surgery. Four patientswith tumour stages II and IIIa remain free of disease.Five out of six patients with tumour progression werein the lowest vaccine dose level of 2.5 µg, while thethree of four patients remaining disease free werevaccinated with 25 µg and 100 µg vaccine doses.

The vaccination did not induce increased frequenciesof CD8+ T-cells reacting with autologous monocytes.HSPPC-96 vaccination also did not lead to increasedfrequencies of patient CD8+ T-cells reactive with



certain defined tumour peptide antigens from CEA,p53 or HeR2/neu. However, an increase of NK-likecytotoxicity could be observed in two of the fourpatients with no evidence of disease. Clonalexpansions of CD8+ T-cells taken from patients duringvaccination are currently being analysed by T-cellreceptor (TCR)-Vβ-spectrotyping.

6.5 Melanoma

In a Phase I study, 36 patients were enrolled afterresection of a > 3cm melanoma metastasis [41]. Ofthese patients, 26 were graded stage IV and 10 weregraded stage III, 30 patients had had prior systemictherapy and 16 had indicator lesions. HSPPC-96 waspurified in sufficient quantities for vaccination fromsamples weighing as little as 2 mg. Patients were givenid. vaccination at doses of 2.5, 25 or 100 µg weekly forfour weeks. Patients showed no serious toxicities totherapy. As of April 2000, one patient at each doselevel had stabilisation or mixed response after initialprogression in nodal or lung metastases, all lastingover seven months. Eleven of the stage IV adjuvantpatients (eight of whom were stage IVM1B) remaineddisease free a median of 11+ months. With 9 monthmedian follow-up, 19 patients remain on study with29 (80%) alive. In vitro studies to assess antitumourimmunity elicited by HSPPC-96 vaccination in thesepatients are ongoing. Previous in vitro data haveshown that HSP70 purified from human melanomacells was able to induce MHC class I-restricted antime-lanoma CTL act ivi ty when presented toMHC-restricted APCs, offering supportive evidencefor a potential in vivo CTL response [42].

7. Expert opinion

HSPPC-96 is truly a personalised tumour vaccine.Being distinct from the traditional chemotherapeuticagents, HSPPC-96 is made from a patient’s own cancerand is generally not in itself cytotoxic. The underlyingmechanism is to stimulate antitumour immuneresponses against the tumour-specific peptideschaperoned by the ubiquitous Hsp, gp96. Preclinicalstudies have shown that HSPPC-96 is broadly effectiveagainst a variety of tumours. The clinical developmentof this unprecedented vaccine has met with severalunique challenges. First, can personalised vaccine bereliably produced and delivered? The answer to thisquestion is certainly yes. The manufacturing ofHSPPC-96 is highly reproducible, predictable and caneventually be automated. Second, what are the best

surrogate immunological markers to determinesuccessful HSPPC-96 immunisation [34]? There are noanswers to this question. Clearly, measuring antitu-mour T-cell response is not enough, as the functionsof APCs and NK cell components are also importantfor the therapeutic effect of HSPPC-96. What are thecontributions of antibody, or B-cell responses? Whatare the roles of other components, such as comple-ment, phagocytic cells and systemic or localcytokines? To summarise, without a proper surrogatemarker, the efficacy of HSPPC-96 can only be judgedby clinical response. This also needs to be betterdefined, since HSPPC-96 itself is not directly cytotoxic.Fourth, how should the clinical response withHSPPC-96 vaccination be evaluated? It is clear fromanimal studies and many of the ongoing clinical trialsthat the immune responses stimulated by HSPPC-96are blunting tumour growth but tumours do notgenerally melt away. The hope, therefore, is toconvert cancer into a manageable chronic disease,which requires new rules and regulations and newways of thinking in order to measure the success of abiological therapy such as immunotherapy withHSPPC-96. Finally, how can immunotherapy withHSPPC-96 be incorporated into existing cancer strate-gies, such as chemotherapy and radiation therapy?Both chemotherapeutic agents and radiation areknown to cause bone marrow suppression. How,then, can immunological competence of the patientbe ensured? How much immunosuppression is ‘safe’and at what level does it become prohibitive for activeimmunisation with HSPPC-96? These questions awaitfurther research.

Nevertheless, the principles learned in rodents aboutthe immunological features of gp96 are so far holdingfirmly in other animals ranging from xenopus tohumans [38]. The gp96 molecule itself is among themost conserved molecules in these animals. It istherefore certain that HSPPC-96 will find a way intothe list of emerging arsenals in the fight against humanmalignancy [40].

Bibliography

Papers of special note have been highlighted as:• of interest•• of considerable interest

1. TRICHOPOULOS D, LIPWORTH L, PETRIDOU E, ADAMIH-O: Epidemiology of cancer. In: Cancer Principles andPractice of Oncology (Edition 5). DeVita VT Jr., Hellman S,Rosenberg SA (Eds.), J B Lipincott-Raven, Philadelphia(1997):231-258.


Caudill & Li 545

2. GILBOA E: The makings of a tumor rejection antigen.Immunity (1999) 11(3):263-270.

3. SRIVASTAVA PK, DELEO AB, OLD LJ: Tumor-rejectionantigens of chemically induced sarcomas of inbredmice. Proc. Natl. Acad. Sci. USA (1986) 83:3407-3411.

4. SRIVASTAVA PK, OLD LJ: Individually distincttransplantation antigens of chemically induced mousetumors. Immunol. Today (1988) 9(3):78-83.

5. LI Z, SRIVASTAVA PK: Tumor rejection antigengp96/grp94 is an ATPase: implications for proteinfolding and antigen presentation. EMBO J (1993)12:3143-3151.

• This article provides the experimental evidence for thepeptide binding properties of gp96.

6. UDONO H, SRIVASTAVA PK: Heat shock protein70-associated peptides elicit specific cancer immunity.J. Exp. Med. (1993) 178:1391-1396.

• An article which demonstrated that the specific immunityinduced by HSP70 is due to the peptides carried by thismolecule.

7. CSERMELY P, SCHNAIDER T, SOTI C, PROHASZKA Z,NARDAI G: The 90-kDa molecular chaperone family:structure, function and clinical applications. Acomprehensive review. Pharmacol. Ther. (1998)79(2):129-68.

8. PARSELL DA, LINQUIST S: The function of heat shockproteins in stress tolerance: degradation and reactiva-tion of damaged proteins. Ann. Rev. Genetics (1993)27:437-496.

9. NIELAND TJ, TAN MC, MONNE-VAN MM, KONING F,KRUISBEEK AM, VAN BLEEK GM: Isolation of animmunodominant viral peptide that is endogenouslybound to the stress protein gp96/grp94. Proc. Natl.Acad. Med. Sci. USA (1996) 93:6135-6139.

10. ISHII T, UDONO H, YAMANO T et al.: Isolation of MHCclass I-restricted tumor antigen peptide and its precur-sors associated with heat shock proteins hsp70, hsp90and gp96. J. Immunol. (1999) 162(3):1303-1309.

11. ARNOLD D, FAATH S, RAMMENSEE HG, SCHILD H:Cross-priming of minor histocompatibility antigen-specific cytotoxic T cells upon immunization with theheat shock protein gp96. J. Exp. Med. (1995) 182:885-889.

•• This article provides evidence for the ability of Hsps tocross-prime antigen-specific CD8+ T cells.

12. BRELOER M, MARTI T, FLEISCHER B, VON BONIN A:Isolation of processed, H-2Kb-binding ovalbumin-derived peptides associated with the stress proteinsHSP70 and gp96. Eur. J. Immunol. (1998) 28(3):1016-1021.

13. BLACHERE NE, LI Z, CHANDAWARKAR RY et al.: Heatshock protein-peptide complexes, reconstituted invitro, elicit peptide-specific cytotoxic T lymphocyteresponse and tumor immunity. J. Exp. Med. (1997)186(8):1315-1322.

14. LINDEROTH NA, POPOWICZ A, SASTRY S: Identificationof the peptide-binding site in the heat shockchaperone/tumor rejection antigen gp96 (Grp94). J.Biol. Chem. (2000) 275(8):5472-5477.

15. LI Z, SRIVASTAVA PK: A critical contemplation on therole of heat shock proteins in transfer of antigenicpeptides during antigen presentation. Behring. Inst.Mitt. (1994) 94:37-47.

16. SUTO R, SRIVASTAVA PK: A mechanism for the specificimmunogenicity of heat shock protein-chaperonedpeptides. Science (1995) 269(5230):1585-1588.

•• This article provides evidence that peptides delivered by theHsp-peptide complex can be cross-presented by APCs.

17. UDONO H, LEVEY DL, SRIVASTAVA PK: Cellular require-ments for tumor-specific immunity elicited by heatshock proteins: tumor rejection antigen gp96 primesCD8+ T cells in vivo. Proc. Natl. Acad. Sci. USA (1994)91(8):3077-3081.

18. SRIVASTAVA PK, UDONO H, BLACHERE NE, LI Z: Heatshock proteins transfer peptides during antigenprocessing and CTL priming. Immunogenetics (1994)39(2):93-98.

19. BASU S, BINDER RJ, SUTO R, ANDERSON KM, SRIVAS-TAVA PK: Necrotic but not apoptotic cell death releasesheat shock proteins, which deliver a partial matura-tion signal to dendritic cells and activate theNF-kappaB pathway. Int . Immunol . (2000)12(11):1539-1546.

20. SINGH-JASUJA H, SCHERER HU, HILF N et al.: The heatshock protein gp96 induces maturation of dendriticcells and down-regulation of its receptor. Eur. J.Immunol. (2000) 30(8):2211-2215.

21. BINDER RJ, HARRIS ML, MENORET A, SRIVASTAVA PK:Saturation, competition and specificity in interactionof heat shock proteins (hsp) gp96, hsp90 and hsp70with CD11b+ cells. J. Immunol. (2000) 165(5):2582-2587.

22. BINDER RJ, HAN DK, SRIVASTAVA PK: CD91 is a receptorfor heat shock protein gp96. Nature Immunol. (2000)1(2):151-155.

•• Article reports the identification of a receptor for gp96.

23. PANJWANI NN, POPOVA L, FEBBRARIO M, SRIVASTAVAPK: The CD36 scavenger receptor as a receptor forgp96. Cell Stress Chaperone. (2000) 5(4):391.

24. SUZUE K, ZHOU X, EISEN HN, YOUNG RA: Heat shockfusion proteins as vehicles for antigen delivery intothe major histocompatibility complex class I presenta-tion pathway. Proc. Natl. Acad. Sci. USA (1997)94(24):13146-13151.

• This article shows that covalent association of Hsps withpeptides can also be used as an effective T-cell vaccine.

25. HUANG Q, RICHMOND JF, SUZUE K, EISEN HN, YOUNGRA: In vivo cytotoxic T lymphocyte elicitation bymycobacterial heat shock protein 70 fusion proteinsmaps to a discrete domain and is CD4(+) T cellindependent. J. Exp. Med. (2000) 191(2):403-408.

26. BEVAN MJ: Minor H antigens introduced on H-2different stimulating cells cross-react at the cytotoxicT cell level during in vivo priming. J. Immunol. (1976)117(6):2233-2238.

27. SIGAL LJ, CROTTY S, ANDINO R, ROCK KL: CytotoxicT-cell immunity to virus-infected non-haematopoietic



cells requires presentation of exogenous antigen.Nature (1999) 398(6722):77-80.

28. SRIVASTAVA PK, MENORET A, BASU S, BINDER RJ,MCQUADE KL: Heat shock proteins come of age:primitive functions acquire new roles in an adaptiveworld. Immunity (1998) 8:656-665.

•• A comprehensive review on immunological properties ofHsps to date.

29. CHANDAWARKAR RY, WAGH MS, SRIVASTAVA PK: Thedual nature of specific immunological activity oftumor-derived gp96 preparations. J. Exp. Med. (1999)189(9):1437-1442.

30. HILF N, SINGH-JASUJA H, SCHERER U, GOUTTEFANGEASC, RAMMENSEE HG, SCHILD H: The heat shock proteingp96 binds specifically to human platelets. Cell StressChaperone (2000) 5(4):387.

31. MENORET A, CHANDAWARKAR R: Heat-shock protein-based anticancer immunotherapy: an idea whose timehas come. Semin. Oncol. (1998) 25(6):654-660.

32. AMATO RJ, MURRAY L, WOOD L, CHERYLYN S,TOMASOVIC S, REITSMA D: Active specific immuno-therapy in patients with renal cell carcinoma (RCC)using autologous tumor derived heat shock protein-peptide complex-gp96 (HSPPC-96) vaccine. Proceed-ings of the 35th Annual ASCO Meeting, Atlanta, USA(1999):1278a.

33. HEIKE M, HERTKORN C, REITSMA DJ et al.: A Phase I trialof heat shock protein - gp96 peptide complex-96(HSPPC-96) as autologous tumor vaccine after surgeryfor gastric cancer. Cell Stress Chaperone. (2000) 5(4):376.

34. SRIVASTAVA PK: Immunotherapy of human cancer:lessons from mice. Nature Immunol. (2000) 1(5):363-366.

35. JANETZKI S, PALLA D, ROSENHAUER V, LOCHS H, LEWISJJ, SRIVASTAVA PK: Immunization of cancer patientswith autologous cancer-derived heat shock proteingp96 preparations: a pilot study. Int. J. Cancer (2000)88(2):232-238.

•• A report of the clinical and immunological results of the firsthuman clinical trial utilising HSPPC-96.

36. AMATO RJ, SAVARY C, TOMASOVIC S, REITSMA D,HAWKINS E, SRIVASTAVA PK: Active specific immuno-therapy in patients with renal cell carcinoma usingautologous tumor derived heat shock protein peptidecomplex 96 vaccine. Cell Stress Chaperone 5(4):373.

37. AMATO RJ, WOOD LS, SAVARY C et al.: Patients withrenal cell carcinoma (RCC) using autologous tumor-derived heat shock protein-peptide complex(HSPPC-96) with or without interleukin-2 (IL-2).Proceedings of the 36th Annual ASCO Meeting, Atlanta, USA(2000):1782a.

38. LEWIS JL, JANETZKI S, LIVINGSTON PO et al.: Pilot trial ofvaccination with autologous tumor-derived gp96 heatshock protein-peptide complex (HSPPC-96) inpatients with resected pancreatic adenocarcinoma.Proceedings of the 35th Annual ASCO Meeting, Atlanta, USA(1999):1687a.

39. ROBERT J, MENORET A, COHEN N: Cell surface expres-sion of the endoplasmic reticular heat shock proteingp96 is phylogenetically conserved. J. Immunol. (1999)163(8):4133-4139.

40. LI Z: Priming of T cells by heat shock protein-peptidecomplexes as the basis of tumor vaccines. Semin.Immunol. (1997) 9(5):315-322.

41. ETON O, EAST M, ROSS MI et al.: Autologous tumor-derived heat-shock protein peptide complex-96(HSPPC-96) in patients (PTS) with metastaticmelanoma. Proc. of the AACR (2000):3463a.

42. CASTELLI C, CIUPITU AT, RINI F et al.: Human heat shockprotein 70 peptide complexes specifically activateantimelanoma T cells. Cancer Res. (2001) 61:222-227.

Marissa M Caudill & Zihai Li†† Author for correspondenceCenter for Immunotherapy of Cancer and Infectious Diseases, MC1601, University of Connecticut School of Medicine, 263Farmington Avenue, Farmington, CT 06030 1601, USATel.: +1 860 679 7979; Fax: +1 860 679 1265;E-mail: [email protected]


Caudill & Li 547

hsppc-96: a personalised cancer vaccine

Documents