Abstract
Heat shock proteins have been the focus of many experimental studies during the last few years in order to understand their biology and their imunologic features. We conducted pre-clinical experiments showing that gp96 purified from human melanoma lines can represent melanoma antigens and stimulate T cells known to recognize such antigens. Clinical studies of vaccination were then initiated by our group by using heat-shock protein gp96 purified from autologous tumor tissues in patients with melanoma and colorectal carcinoma. The results of these trials in metastatic melanoma patients with measurable disease showed that a melanoma-specific T cell response can be generated or increased in approximately 50% of vaccinated patients. Moreover, signs of clinical responses were obtained consisting of two complete responses and three long-lasting stabilizations. Similar results were obtained in patients with liver metastases of colorectal cancer made disease-free by surgery. In both studies a clear association was found between T cell immune response induced by the vaccine and clinical response both in the trial of melanoma (tumor response) and in that of colorectal cancer patients (disease-free and overall survival at 5 years).
Introduction
During the last few years heat shock proteins (HSPs) have attracted the attention of many investigators in the field of tumour immunology and immunotherapy thanks to the demonstration that some of these proteins display important immunobiological functions. Among such HSP families, HSP 70 and GRP34 (hereafter gp96) have been the focus of investigation and the role of gp96 as chaperone of several tumour antigen peptides was demonstrated in mice and humans, leading to the use of these proteins as immunogens in animal models and cancer patients Citation[1], Citation[2].
In vitro and ex-vivo immunogenicity of gp96 in the human system
Gp96 obtained from diverse murine tumours were shown to induce an individually tumour-specific T-cell response that could protect from the subsequent challenge with the tumour from which gp96 were derived but not against challenge with unrelated neoplasms Citation[1]. The mechanism by which gp96 can chaperone immunogenic peptides is unclear Citation[3], Citation[4], particularly in human cells, but could be mediated by binding of the gp96/peptide complexes to specific receptors of the antigen presenting cells (APCs), internalization of the complex and re-presentation of epitopes to T-cells Citation[1], Citation[4].
This study tested whether re-presentation of known tumour antigens by APCs pulsed with autologous tumour-derived gp96 may occur by using patients’ monocytes or dendritic cells as APCs. Peripheral blood mononuclear cells (PBMCs) were obtained from patients with either melanoma or colorectal cancer. Monocytes separated from such PBMCs were pulsed in vitro with 10 µg of purified, tumour-derived gp96 or with 5 µM of peptide and used to stimulate either CD8+T cell clones known to recognize melanoma antigens like gp100 or Melan-A/MART1 or colorectal cancer antigens like Ep-CAM or CEA. These experiments showed that APC pulsed with tumour-derived gp96 were recognized by antigen-specific T-cells in a HLA class I-restricted fashion as evaluated by the release of IFN-γ in different assays including ELISPOT and cytokine intra-cellular staining Citation[5]. Similar results were reported by other groups Citation[6], Citation[7]. By using the HLA/epitope tetramer staining technique, one could also enumerate T-cells recognizing such epitopes in the blood of cancer patients, showing that a spontaneous, though weak immune response to self-antigens (e.g. CEA, Melan-A/MART1) may potentially occur in vivo without prior stimulation in at least a fraction of cancer patients Citation[5].
Therefore, based on this series of in vitro experiments and on previous studies in animal models, a phase I–II clinical trial was initiated to assess the tolerability and the immunogenicity of the administration of autologous tumour-derived gp96 in metastatic melanoma patients.
Clinical studies with gp96
Melanoma
The first trial included the administration of autologous metastatic tumour-derived gp96 (also defined as Heat Shock Protein Peptide Complex or HSPPc-96, Oncophage®) purified at the Antigenics facility, to melanoma patients with metastatic disease or made disease-free by surgery. The vaccine included gp96 as defined by a Western blotting assay. Patients received 5 or 50 µg of the protein subcutaneously four times, 1 week apart (1st cycle), followed by a 4-week rest and then a subsequent cycle of four injections given 2 weeks apart Citation[8]. Patients were carefully monitored for appearance of adverse events, for the development of T-cell-mediated immune response by ELISPOT and tetramer staining and for tumour response according to the RECIST criteria. No major (grade 3 or 4) local or systemic toxicity was observed while 50% of patients developed a melanoma antigen (Melan-A/MART1) or autologous melanoma cell-specific T-cell reaction in the PBMCs which was sustained during the vaccination period Citation[8]. Important signs of clinical response were also recorded which included two complete and durable responses (2 and 6 years, the latter still ongoing; the first one then re-assessed as partial response) and three long-term stable disease in the 28 assessable patients.
Thus, this study indicated that vaccination with autologous tumour-derived gp96: (a) is well tolerated, (b) shows a melanoma antigen-specific T-cell response can be elicited in metastatic patients, and (c) shows clear sign of potential therapeutic activity.
A more recent trial aimed at increasing the frequency of the immune response and, hopefully, of the clinical response by adding to the Oncophage® vaccine the administration of GM-CSF and of IFN-α, the first cytokine given locally at the site of vaccination to increase the recruitment of APCs Citation[9] and the second one systemically in an attempt to improve the expression of class I HLA often down-regulated in metastatic melanoma cells Citation[10]. Results of this study, while confirming the immunogenicity of melanoma-derived gp96, was rather disappointing since both the frequency of patients showing the anti-melanoma T-cell response in PBMCs (32%) and that of clinical responses (11 stable diseases out of 20 assessable patients) were lower than that observed in the previous study Citation[8] (). A possible reason for such finding may lie in the increase in myeloid suppressor cells that was found in the PBMCs of melanoma patients after the local administration of GM-CSF (G. Parmiani, L. Rivoltini, unpublished data).
Table I. Increased frequency of anti-tumour T and NK cells in patients vaccinated with autologous tumour-derived HSPPC-96.*
Colorectal cancer
That gp96 vaccines may be immunogenic in other tumours in addition to melanoma was tested in an additional clinical study performed in patients with resectable liver metastases of colorectal cancer. In this phase I–II trial 29 patients were made disease-free after surgical resection of liver metastases of colorectal cancer, the tumour tissue was used to manufacture the gp96-based vaccine and patients treated in an adjuvant setting with a schedule similar to that mentioned above for melanoma patients. Again more than 50% of subjects developed a T-cell-mediated, class I HLA-restricted immune response against colorectal cancer cells and/or Ep-CAM or CEA HLA-A0201 peptides Citation[11] (). Moreover, patients who showed an immune response had statistically significant prolongation of the disease-free survival and an overall survival which is maintained at 5 years of observation Citation[11].
Immune response and clinical response after gp96 vaccination
A major problem in vaccination studies is that the frequency and strength of the immune response to vaccine is still not predicting the clinical outcome. Therefore, this study analysed the possible relationship between these two parameters in the three trials of vaccination with gp96. The results are summarized in and suggest that patients who developed or increased their anti-tumour-specific immune response after vaccination did much better in terms of clinical response than patients unable to mount the immune response. However, given the limited number of subjects treated, the clinical significance of this association needs to be confirmed in larger clinical trials of gp96.
Table II. Association between clinical response, T-cell response and HLA/melanoma associated antigen (MAA) expression by the original tumour.
Conclusions
The above studies clearly indicate that gp96 can generate a T-lymphocyte anti-tumour-specific immune response in at least half of the patients treated and that this response is associated with a clinical response. However, up to 20% of subjects who did not show clinical response mounted an immune response. In addition, different tumour types can respond to gp96 immunotherapy supporting the further use of such a vaccine strategy in cancer. Some points of weakness though need to be overcome to increase the applicability of this strategy. In fact, no clear and reproducible evidence of involvement of individual, unique tumour antigens in the immune response has been provided so far in patients, leaving the principle on which this strategy is based still not proved in the human system. Moreover, the requirement of 5–7 g of fresh, non-necrotic tumour tissue necessary for manufacturing the vaccine limits substantially the number of patients that could be treated.
Work is ongoing at the pre-clinical and clinical levels to bypass such drawbacks and to optimize the use of gp96 as therapeutic vaccine in cancer patients.
Acknowledgements
The authors’ work was supported by the Italian Association for Cancer Research (Milan) and by Antigenics Inc. (New York).
References
- Srivastava PK. Interaction of heat shock proteins with peptides and antigen presenting cells: Chaperoning of the innate and adaptive immune responses. Annu Rev Immunol 2002; 20: 395–425
- Parmiani G, Testori A, Maio M, Castelli C, Rivoltini L, Pilla L, Belli F, Mazzaferro V, Coppa J, Patuzzo R, Sertoli MR, Hoos A, Srivastava PK, Santinami M. Heat shock proteins and their use as anticancer vaccines. Clin Cancer Res 2004; 10: 8132–8141
- Nicchitta CN. Re-evaluating the role of heat-shock protein-peptide interactions in tumour immunity. Nature Rev Immunol 2003; 3: 427–432
- Binder RJ, Srivastava PK. Peptides chaperoned by heat-shock proteins are a necessary and sufficient source of antigen in the cross-priming of CD8+cells. Nature Immunol 2005; 6: 593–599
- Rivoltini L, Castelli C, Carrabba MA, Mazzaferro V, Pilla L, Huber V, Coppa J, Gallino G, Scheibenbogen C, Squarcina P, Cova A, Camerini R, Lewis JJ, Srivastava PK, Parmiani G. Human tumor-derived heat shock proteins 96 mediates in vitro activation and in vivo expansion of melanoma and colon carcinoma-specific T cells. J Immunol 2003; 171: 3467–3474
- Noessner E, Gastpar R, Milani V, Brandl A, Hutzler PJS, Kuppner MC, Roos M, Kremmer E, Asea A, Calderwood SK, Issels RD. Tumor-derived heat shock protein 70 peptide complexes are cross-presented by human dendritic cells. J Immunol 2002; 169: 5424–5432
- Staib F, Distler M, Bethke K, Schmitt U, Galle PR, Heike M. Cross-presentation of human melanoma peptide antigen MART-1 to CTLs from in vitro reconstituted gp96/MART-1 complexes. Cancer Immun 2004; 4: 3
- Belli F, Testori A, Rivoltini L, Maio M, Andreola S, Sertoli MR, Gallino G, Piris A, Cattelan A, Lazzari I, Carrabba M, Scita G, Santantonio C, Pilla L, Tragni G, Lombardo C, Arienti F, Marchiano A, Queirolo P, Bertolini F, Cova A, Lamaj E, Ascani L, Camerini R, Corsi M, Cascinelli N, Lewis JJ, Srivastava P, Parmiani G. Vaccination of metastatic patients with autologous tumor-derived heat shock proteins gp96. J Clin Oncol 2002; 20: 4169–4180
- Dranoff G, Jaffee E, Lazenby A, Golumbek P, Levitsky H, Brose K, Jackson V, Hamada H, Pardoll D, Mulligan RC. Vaccination with irradiated tumor cells engineered to secrete murine granulocyte-macrophage colony-stimulating factor stimulates potent, specific, and long-lasting antitumor imunity. Proc Nat Acad Sci USA 1993; 90: 3539–3543
- Pilla L, Patuzzo R, Rivoltini L, Maio M, Pennacchioli E, Lamaj E, Maurichi A, Massarut S, Marchiano A, Santatonio C, Tosi D, Arienti F, Cova A, Sovena G, Pins A, Nonaka D, Bersani I, Florio AD, Luigi M, Srivastava PK, Hoos A, Santanami M, Parmiani G. A phase II trial of vaccination with autologous, tumor-derived heat shock protein peptide complex Gp96, in combination with GM-CSF and interferon α in metastatic melanoma patients. Cancer Immunol Immunother. published online. 2005
- Mazzaferro V, Coppa J, Carrabba MA, Rivoltini L, Schiavo M, Regalia F, Mariani L, Camarini T, Marchianò A, Andreola S, Camerini R, Corsi M, Lewis JJ, Srivastava PK, Parmiani G. Vaccination with autologous tumor-derived heat shock proteins gp96 after liver resection for metastatic colon cancer. Clin Cancer Res 2003; 9: 3235–3245