This Article Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Chapman, P. B. Search for Related Content PubMed PubMed Citation Articles by Chapman, P. B. Social Bookmarking                            What's this?

Vaccinating Patients With Autologous Tumor

Memorial Sloan-Kettering Cancer Center, New York, NY

PHYSICIANS TAKING care of cancer patients have long tried to harness the power of the immune system to kill cancer cells. The earliest attempts were reported a hundred years ago at a time when few of the basic rules of immunity were known.¹ Not surprisingly, the early empiric attempts to immunize against cancer used the same approach used to immunize against infectious agents. Patients were inoculated with crude preparations of the tumor—either the patient’s own tumor or someone else’s—in an era when it was not possible to analyze immune responses and when careful clinical trials were typically not conducted. As a result, these early attempts largely cannot be interpreted and the field could not advance.

Over the past two decades, there has been an explosion of knowledge regarding the antigens expressed on tumor cells, antigen presentation, mechanisms of antibody and cellular recognition of antigens, and how tumor cells can avoid being recognized by the immune system. The challenge has been to incorporate this continually advancing understanding into new vaccine strategies against cancer. At this point, cancer vaccines can be divided into two broad categories: vaccines formulated with purified, defined antigens and vaccines made from undefined antigen mixtures derived from tumor cells. This latter type of vaccine has two potential advantages. First, these vaccines are not restricted merely to the antigens we know about and can purify. Rather, these vaccines may contain hundreds or thousands of potentially antigenic proteins, carbohydrates, and lipid molecules. Second, in the case of autologous vaccines, there is the possibility of containing antigens that are unique to the tumor cell and not expressed on normal tissues—such as tumor-specific mutations or translocations—that may as yet be unknown to us.

On the other hand, undefined vaccines also have some drawbacks. Although there may be thousands of different molecules in the vaccine preparation, the vast majority of them are not likely to be relevant immunogens because they are either common self-molecules expressed on multiple normal tissues (eg, actin) or, in the case of allogeneic vaccines, alloantigens not expressed on the patient’s tumor cells (eg, HLA molecules or other polymorphisms). These alloantigens are likely to be the most immunogenic molecules in the mixture. Because there are so many potential antigens present, and because we do not know which one(s) might be responsible for tumor rejection, it is often difficult to devise relevant in vitro immunologic assays to monitor the induced immune response. Without an immune response to guide development of these vaccines, clinical responses—such as tumor shrinkage—become the only indicator of efficacy. This represents a very stringent test to guide vaccine development and is only possible in the patient with advanced cancer. Animal models suggest that vaccines may be more effective in minimal disease settings, leading many investigators to test vaccines in patients clinically free of tumor after surgery. However, the clinical end point in these trials is progression-free survival or overall survival, requiring a randomized trial comparing vaccine to a control group. A third drawback with undefined vaccines is that it is hard to know for sure that each batch is the same. Indeed, for vaccines made from the patient’s own tumor, each batch is unique.

In this issue of the Journal of Clinical Oncology, Belli et al² have used an undefined autologous vaccine approach in melanoma patients in order to take advantage of the recent understanding of the role heat-shock proteins (HSPs) play in antigen presentation. HSPs are highly conserved proteins that serve as chaperones for newly synthesized proteins. One of their roles seems to be to complex with peptides present in the cytoplasm and guide the peptide into the major histocompatibility complex class I pathway. HSP-peptide complexes presented to antigen-presenting cells seem to be taken up efficiently, resulting in peptide presentation. Mice immunized with HSP-peptide complexes purified from tumor cells can develop striking antitumor immunity. Belli et al report results from the immunization of advanced melanoma patients with HSP-peptide complexes purified from their own melanoma tumor.

The potential advantage of this approach is that the patient is inoculated with a subset of autologous tumor-derived peptides that may be enriched for peptides that can be presented by HLA class I molecules. Further, since HSPs facilitate uptake and presentation of the peptides by antigen-presenting cells, the HSPs function as a kind of immune adjuvant. However, purifying the HSP-peptide complexes is a very labor-intensive approach that requires 3 g of tumor to be collected and transported to the laboratory for HSP purification and vaccine formulation. Of the 64 patients who had tumor collected, 40% could not receive the initial four weekly injections (ie, one injection per week for 4 weeks), mainly because not enough vaccine could be made or because their melanoma progressed while vaccine was being prepared.

This was a pilot trial and was not designed to assess clinical efficacy. As a result, the clinical results must be interpreted with extreme caution. The most intriguing part of the report is that two of the 28 patients treated with measurable disease had complete response in multiple, small soft tissue metastases while receiving the vaccine. This is consistent with other reports of undefined allogeneic^3,4 or autologous^5-7 melanoma vaccines in which occasional clinical responses have also been observed. Another interesting observation, also seen in other trials, is that even when vaccine therapy resulted in death of all melanoma cells in a tumor, the tumor mass itself, made up of necrotic and inflammatory cells, persisted. Some of the patients were free of disease when they started the vaccinations, but it is not possible to know whether the vaccine contributed to the prolonged progression-free periods seen in some of these patients.

Of the 39 patients vaccinated, T-cell responses were measured in 23 patients against either autologous melanoma (17 patients) or allogeneic melanoma cell lines matched for a single HLA-A allele (six patients). Use of allogeneic cell lines is clearly not optimal, since it provides the possibility of confounding alloreactivity while at the same time limiting the ability to observe reactivity to epitopes presented by the patient’s other five HLA molecules. In five of 17 patients, T-cell reactivity increased against autologous melanoma after immunization. The fundamental question is whether detectable immunologic responses correlate with clinical responses. Ultimately, this question remains unanswered in this study. This is because, first, there was only a small number of subjects and immunologic data were not available for a substantial proportion of the patients. For example, there were T-cell reactivity data only for half of the 28 patients with measurable disease. Second, it is not clear what should constitute a clinical response. Should prolonged stable disease be attributed to the vaccine? I believe this is perilous. If we consider only objective responders (ie, the two patients with complete responses), there does not seem to be a correlation between T-cell response and clinical response among the patients with measurable disease.

One mechanism by which tumor cells might avoid being killed by T cells is by loss of HLA expression. Belli et al² tested 20 of the melanoma tumors for HLA expression by immunohistochemistry using a monoclonal antibody that recognizes all class I HLA gene products. This means that a tumor must lose expression of HLA-A, -B, and -C alleles in order to stain negative for this antibody. They found that 40% of tumors tested had limited staining by this antibody, which means that all HLA class I alleles were downregulated, which is a sobering observation. As the authors point out, this analysis underestimates the amount of HLA loss, since tumor cells losing individual HLA alleles—but not all six alleles—would still stain as positive with this antibody. Previous studies have shown that individual HLA allele loss is rather common in melanoma,⁸ which could result in selection of clones no longer able to present tumor rejection antigens.

From this article,² we may conclude that this approach to vaccination seems to be safe; there were no serious side effects. Whether it is feasible or not depends on one’s point of view. Given that 40% of patients accrued could not be treated with at least four vaccinations speaks to the technical difficulties involved in producing this vaccine. An increase in T-cell reactivity against autologous melanoma could be detected in approximately one third of the patients tested, although we do not yet know anything about the antigens recognized or whether the reactivity is sufficient to reject tumor. Rare tumor responses were observed, which is encouraging, but it remains unclear whether T-cell reactivity correlates with clinical responses. Deciding whether to try to improve the immunogenicity of the vaccine or to test the vaccine in the adjuvant setting will be important for the future development of this vaccine.

REFERENCES

1. Southam CM: Applications of immunology to clinical cancer: Past attempts and future possibilities. Cancer Res 21: 1302-1316, 1961[Abstract/Free Full Text]

2. Belli F, Testori A, Rivoltini L, et al: Vaccination of metastatic melanoma patients with autologous tumor-derived heat shock protein gp96-peptide complexes: Clinical and immunologic findings. J Clin Oncol 20: 4169-4180, 2002[Abstract/Free Full Text]

3. Mitchell MS, Von Eschen KB: Phase III trial of Melacine melanoma theraccine versus combination chemotherapy in the treatment of stage IV melanoma. Proc Am Soc Clin Oncol 16: 494a, 1997 (abstr 1778)

4. Hsueh EC, Nathanson L, Foshag LJ, et al: Active specific immunotherapy with polyvalent melanoma cell vaccine for patients with in-transit melanoma metastases. Cancer 85: 2160-2169, 1999[CrossRef][Medline]

5. Berd D, Maquire H, McCue P, et al: Treatment of metastatic melanoma with an autologous tumor-cell vaccine: Clinical and immunological results in 64 patients. J Clin Oncol 8: 1858-1867, 1990[Abstract]

6. Berd D: Autologous, hapten-modified vaccine as a treatment for human cancers. Vaccine 19: 2565-2570, 2001[CrossRef][Medline]

7. Soiffer R, Lynch T, Mihm M, et al: Vaccination with irradiated autologous melanoma cells engineered to secrete human granulocyte-macrophage colony-stimulating factor generates potent antitumor immunity in patients with metastatic melanoma. Proc Natl Acad Sci U S A 95: 13141-13146, 1998[Abstract/Free Full Text]

8. Kageshita T, Wang Z, Calorini L, et al: Selective loss of human leukocyte class I allospecificities and staining of melanoma cells by monoclonal antibodies recognizing monomorphic determinants of class I human leukocyte antigens. Cancer Res 53: 3349-3354, 1993[Abstract/Free Full Text]

CiteULike    Complore    Connotea    Del.icio.us    Digg    Facebook    Reddit    Technorati    Twitter    What's this?

This Article Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Chapman, P. B. Search for Related Content PubMed PubMed Citation Articles by Chapman, P. B. Social Bookmarking                            What's this?