Introduction

In the United States, epithelial ovarian carcinoma accounts for 90% of ovarian malignancies and is the second most common gynecologic cancer. More than 25,000 new cases were reported in 1998, and the disease causes an estimated 14,500 deaths annually.¹ The majority of patients are diagnosed when the disease is advanced, and risk factors remain largely unknown. A family history of ovarian cancer and mutations in the BRCA1 or BRCA2 genes are important predisposing factors for familial disease, which accounts for approximately 10% of cases. Because the diagnosis is made at an advanced stage and because ovarian tumors are only moderately sensitive to chemotherapy, cure rates have remained essentially unchanged for many years. Mortality is usually caused by intestinal obstruction resulting from extensive peritoneal and serosal involvement. Treatment involves cytoreductive surgery, including a bilateral salpingo-oophorectomy and in most cases a total hysterectomy, followed by platinum-based chemotherapy with or without paclitaxel. Treatment for recurring or persistent disease after first-line chemotherapy may include reinduction therapy with first-line chemotherapy agents such as carboplatin or the taxanes, high-dose chemotherapy, or second-line agents such as hycamptamine. Further treatment for recurrent or persistent disease is essentially palliative. Therapies under investigation include a variety of novel immunotherapy and gene transfer approaches, among other modalities.

Recent advances in tumor immunology are focusing attention on antigen epitopes, mechanisms for recognition of these epitopes, and specific mechanisms involved in the activation of T cells against tumor cells. In addition, the role of immune suppressor factors and the interactions with other factors, such as those that may promote angiogenesis and peritumoral fibrosis in the tumor microenvironment, are receiving more consideration.

The first generation of therapeutic vaccines against ovarian cancer involved the use of irradiated autologous or allogeneic tumor cells or extracts of tumor cells. Researchers at the University of Texas M.D. Anderson Cancer Center employed lysates of allogeneic tumor cells obtained from tumor cell lines that had been infected with an attenuated form of the influenza virus. This approach could be considered a prototype for the gene transfer therapies of today. Viral infection of tumors in vivo was shown to render animals tumor-free through the induction of viral oncolysis, and animals were also found to be immune to rechallenge with identical tumors. In a different approach, bacillus Calmette-Guérin (BCG) was co-administered with tumor cells with encouraging results.^2,3

Peptides derived from tumor-associated or tumor-specific (mutated) antigens are presented in the grooves of major histocompatibility complex (MHC) class I or class II molecules, where they activate either CD8+ T cells with cytotoxic T-lymphocyte (CTL) activity or CD4+ T lymphocytes that secrete cytokines capable of inhibiting the growth of tumor cells. Dendritic cells may be the most important of the antigen-presenting cells (APCs) since they are able to present antigen-derived peptides and provide the necessary stimuli through co-stimulation (eg, B7 molecules) and through the actions of certain accessory factors (eg, intercellular adhesion molecules [CD54]). It is possible that tumor antigens derived from allogeneic tumor cell vaccines can be reprocessed and presented by APCs such as dendritic cells or certain macrophages. Research conducted at our center has demonstrated the presence of APCs in the peritoneal cavities and in the blood of ovarian cancer patients.⁴ However, the surface antigen characteristics of these cells suggest that further maturation may be required for optimum antigen-presenting function. Current research efforts are examining methods in which dendritic cells can be manipulated so that they are phenotypically and functionally matured.

Tumor-cell peptide epitopes are generally recognized by T lymphocytes in the context of MHC class I or class II molecules, although other mechanisms for T-cell receptor activation do not require presentation in the context of the MHC (eg, MUC1 antigen).

Peptide MHC-mediated activation of T cells is subverted when MHC expression is down-regulated or absent. Co-stimulatory molecules such as B7 and other accessory molecules such as intercellular adhesion molecules (ICAM) need to be co-presented with tumor antigen peptide-loaded MHC molecules in order to simulate an effective antitumor cell response by the T cells.⁵ It is thought that the use of recombinant cytokines such as recombinant interleukin-2 (rIL-2), which itself stimulates cytotoxic lymphocytes belonging to the T cell or natural killer (NK) cell lineages, might replace the need for co-stimulation. However, co-stimulatory antigens have other important functions, including protection of activated lymphocytes from apoptosis. Moreover, rIL-2 may have pleiotropic functions that could also interfere with T cell activation. The NK immune system is phylogenetically more primitive and may have a more important role in the control of hematologic malignancies than in the control of solid malignancies in vivo. Moreover, NK cells are present in much smaller numbers than T cells in association with ovarian cancer.

Cultures of mixtures of ovarian tumor cells and T lymphocytes have been shown to generate either CD8+TCR-alpha-beta+ or CD4+TCR-alpha-beta+ T-cell lines in vitro.⁶ The CD8+TCR-alpha-beta+ T-cell lines exhibited preferential cytotoxicity against autologous ovarian tumor cells in the presence of low concentrations of rIL-2^7,8 or produced interferon-gamma (IFN-gamma), granulocyte-macrophage colony-stimulating factor (GM-CSF) and tumor necrosis factor-alpha (TNF-alpha) in an antigen-dependent or antigen-independent fashion.⁹ Although adoptive therapy using these T-cell lines has been employed by us and by others, further research is needed to improve the success rate for the expansion of these T-cell lines. The lines are, however, useful for identifying antigen epitopes or peptides that could be employed as tumor vaccines.

Cytokines are believed to attract lymphocytes to the site of malignancy by stimulating the local production of chemokines. Cytokines such as IFN-gamma, IL-2, TNF, and GM-CSF contribute to the proliferation and differentiation of mononuclear lymphocytes, including monocytes, macrophages, B cells, and dendritic cells, that are involved with antitumor immunity. Recombinant IL-2 enhances the cytotoxic activity of T cells and non-T cells,^10,11 but it also can stimulate other cells, including cells of the monocyte/macrophage lineage. Like many other cytokines, rIL-2 is pleiotropic and thus can stimulate other cells or the production of other cytokines whose effects differ from that of rIL-2 and may even suppress T-cell activation.¹² Recombinant IL-12, which was previously called NK cell-stimulating factor, is the largest cytokine described to date. It comprises two subunits of 35 and 40 kDa that stimulate cytotoxic activity of NK cells and T lymphocytes, respectively.¹³ Recombinant IL-12 binds to its beta-2 receptor on activated T cells, which results in the secretion of IFN-gamma, a function of T_H1 cells. In contrast, T_H2 cells secrete either B-cell stimulatory cytokines (eg, IL-4, -5, and -6) or suppressor cytokines such as IL-10. Several interferon-inducible proteins have been detected following the activation of T_H1 cells by rIL-12. One of these is interferon-inducible protein-10, a C-X-C chemokine. This chemokine has important biologic effects, including an antitumor response mediated through the down-regulation of angiogenesis-inducing molecules such as basic fibroblast growth factor.¹⁴ The response of T cells to rIL-12 is increased if T_H1 cells have been previously activated. A recent report from our group on the intraperitoneal injection of rIL-12 in patients with ovarian carcinoma is showing clinical activity without significant drug-related toxicity.¹⁵

Cytokines can induce a strong adaptive immune response but can also produce inhibitory effects. Intraperitoneal rIL-2 administration has resulted in increased concentrations of IL-10 in the peritoneal cavities of patients with ovarian cancer.⁶ Recent studies in our laboratory suggest that CD14+DR– monocytes are largely responsible for the production of IL-10 in ovarian cancer patients, whereas both monocytes and tumor cells can contribute to the production of transforming growth factor-beta (TGF-beta) isotypes, which in their activated state are also potent immune-suppressor molecules.¹⁶

Tumor vaccines may be employed therapeutically or for prophylaxis after primary therapy. Vaccines that enhance or generate humoral responses produce antibodies that can be detected over a relatively long period. To be effective, these antibodies need to be capable of targeting cell surface antigens in live cell assays. Maintaining specific cellular immune responses to antigen epitopes (adaptive immunity) may require more frequent immunizations, although memory cells can sustain the ability to respond to rechallenge with the immunizing epitope. Again, the epitope has to be recognized by specifically sensitized CTL on the tumor cells that express these antigens in their natural state. Several vaccine approaches have been employed, including virus-augmented vaccines, peptides derived from tumor-associated antigens, intact irradiated tumor cells transduced with co-stimulatory or other antigens, and carbohydrate or glycolipid vaccines. Recently, DNA vaccine approaches have been introduced.

First-Generation Vaccines

Peptide Vaccines

Peptide vaccines require some form of adjuvant to enhance the immune response. The optimum peptide-based vaccine may need to consist of multiple peptides (polyvalent) to account for both intratumoral heterogeneity and haplotype differences. Antigens that are restricted to HLA-A2+ cells will likely be ineffective in more than 50% of patients of any ethnic group, and in fact, the frequency of the A2+ haplotype in ovarian cancer patients is unknown. The number of known MHC motif-matching peptides from ovarian cancer is small, and it will be necessary to develop polyvalent vaccines that elicit CTL or activate T cells that secrete tumor-inhibitor cytokines.

We have also studied the effects of a sialylated Tn (sTn) antigen vaccine in a recent clinical trial.²⁷ Carbohydrate-derived antigens not expressed on normal cells include O-linked mucin glycans Tn, T, and sTn. The clinical trial used a vaccine that incorporated a synthetic analog of sTn [NANAalpha (2-6) Gal-NAc] conjugated with keyhole limpet hemocyanin (KLH) and mixed with Detox-B SE adjuvant (Biomira Inc, Edmonton, Alberta, Canada) for injection into patients. The vaccine was well tolerated and produced an immunologic response in vivo. Although no objective responses were observed at either of two doses examined, a number of patients had stable disease for prolonged periods.²⁷

To be effective, vaccines for patients with ovarian cancer will need to comprise a broad range of epitopes to compensate for tumor heterogeneity. In the case of carbohydrate or glycolipid antigens, it is essential to be able to demonstrate antibody binding to cell-surface antigens in live cell assays using the serum of immunized patients. Carbohydrate-derived vaccines also require co-administration of adjuvants, and the safety of these adjuvants will need to be examined when different routes of administration are used.

Polynucleotide Vaccines

Genetic immunization is a relatively new process that uses polynucleotides as vaccines made up of DNA to promote the stimulation of an immune response.²⁸ DNA incorporated in a bacterial plasmid and containing a strong eukaryotic promoter to ensure transcription of a target gene is inoculated into the cells of interest, and the expression of plasmid-encoded protein allows the host to generate an adequate immune response. Protective immune responses have resulted from using DNA-encoded viral, bacterial, parasitic, and tumor antigens.²⁹ Some advantages of these vaccines include their ease of construction and ability to produce a long-lasting immune response. These responses have induced lifelong protective immunity in mice.³⁰

Plasmid-encoded antigen is processed similarly to antigen synthesized from virus and results in activation of CTL and stimulation of T-helper cells and B cells. This type of vaccine is more stable and less expensive than most recombinant protein vaccines. It can be directly injected into cells by syringe or coated on microgold particles for injection by a gene gun. It also can be inserted into the genome of viruses that are used as vectors for insertion. These methods have been shown to be both effective and safe in neonatal mice³¹ and could have potential value in patients with ovarian cancer.

Vectors that encode for genes, cytokines, or co-stimulatory molecules can be employed to induce antitumor immune responses in vivo. Studies in our laboratory²¹ are examining the effects of using autologous ovarian tumor cells that have been transduced with the gene that encodes for the naturally occurring B7.1 (CD80) co-stimulatory molecule (Fig 2). The low expression of the CD80 molecule on peritoneal exudate cells may interfere with activation of an effective immune response in vivo. A canarypox vector that incorporates the gene for hB7.1 (ALVAC-hB7.1) (Pasteur Merieux Connaught, Strasbourg, France) has been constructed and safely administered to healthy subjects. We are producing autologous tumor cell vaccines that include tumor cells obtained from patients during surgery for ovarian cancer or from malignant ascites. The tumor cells are infected with ALVAC-hB7.1 after treatment of the cells in vitro with recombinant IFN-gamma (rIFN-gamma). This results in the production of an autologous tumor cell vaccine that exhibits increased expression of HLA class I and class II, CD54 (ICAM), and the CD80 (B7.1) co-stimulatory molecule. To inhibit their growth, the tumor cells are irradiated before readministration to patients. This tumor vaccine requires a vector that can efficiently infect cells and transduce the gene of interest. It is hypothesized that tumor cells that are modified in this manner might perform the functions of APCs. Vaccines of this type could be efficient at activating T cells against tumor-associated antigens that are expressed by the patient’s own tumor and in the context of the patient’s own HLA haplotype. A typical result showing expression of CD80 (B7.1), HLA class I, and HLA class II on autologous ovarian tumor cells after infection of the cells with ALVAC-hB7.1 is shown in Fig 1. A possible limitation to the use of autologous vaccines is the amount of tumor, 50 g or more, that is required for the production of multiple doses. However, such large amounts of tumor and, in certain patients, large number of ascites containing tumor cells are available from patients who are undergoing initial or salvage surgery for ovarian cancer. These vaccines can be prepared and stored at ultra-low temperatures until administered, usually following reduction of the tumor burden with chemotherapy or secondary cytoreductive surgery.

Fig 2. - Antigen expression in tumor cells infected with ALVAC-hB7.1. Freshly obtained tumor cells treated with recombinant interferon-gamma and then infected with the ALVAC-Hb7.1 canarypox vector that expresses the gene for human 7.1. Results obtained by fluorescence-activated cell sorting show high expression of B7.1 (CD80) and increased expression of HLA class I and class II antigens in contrast to almost no expression of B7.1 in the non-infected cells.

How a vaccine is administered needs consideration. Different routes of vaccine administration may stimulate the populations of mononuclear leukocytes differently.³² Vaccines have been administered by the following routes: oral, intranasal, intravenous, intradermal, intramuscular, and intraperitoneal. In the case of ovarian cancer, the routes used are intradermal, intramuscular, and intraperitoneal. There could be a particular advantage to the use of the intraperitoneal approach, which might stimulate immune cells at the site of the tumor, bypassing the capillary barrier.²¹

Intradermal or Intracutaneous Injection

Vaccines injected into the skin encounter professional APCs called Langerhans cells (dendritic cells). Antigens incorporated into vaccines are transported to the draining lymph nodes, where T-cell activation occurs. This route of administration produces systemic immune responses to an antigen. Dosing and optimal frequency of administration need to be determined for each vaccine approach because responses may be either enhanced or decreased with repeated vaccinations. It is unknown whether peripheral sites of vaccination would generate CTL responses of a sufficient degree to target tumor cells that are growing on surfaces in the peritoneal cavity. It is possible that both intracutaneous and intraperitoneal routes might be needed.

Vaccines injected intramuscularly are taken up by APCs (primarily dendritic cells) or the muscle cells. Muscle cells are thought to be the primary cells of infection, but they express low levels of MHC class I molecules and no MHC class II molecules. Vaccines that encode for tumor antigens and that are expressed by muscle cells would therefore have to be taken up by APCs and re-presented (cross-priming in the context of MHC molecules for T-cell activation).³¹ Vaccines that encode genes for cytokines and tumor antigens can also be taken up by APCs for activation of lymphocytes. A potential problem with this route of vaccination, which has also been seen after intradermal or intracutaneous vaccination, is that only a small number of lymphocytes able to produce an effective adaptive immune response are activated. This might be a less important issue for patients who are immunized with vaccines that are designed to generate antibodies in contrast to activated CTLs. Responses may involve innate immunity, and a memory-dependent strong effector CD8+CTL response may not develop following this approach.

Conclusions

References

1. Landis SH, Murray T, Bolden S, et al. Cancer statistics, 1998. CA Cancer J Clin. 1998;48:6-29.

2. Hersh EM, Gutterman JU, Mavligit GM. BCG as adjuvant immunotherapy for neoplasia. Annu Rev Med. 1977;28:489-515.

3. Florentin I, Bruley-Rosset M, Davigny M. Lymphocyte macrophage interactions in BCG-treated mice. Recent Results Cancer Res. 1976;56:49-57.

4. Melichar B, Savary C, Kudelka AP, et al. Lineage-negative human leukocyte antigen-DR+ cells with the phenotype of undifferentiated dendritic cells in patients with carcinoma of the abdomen and pelvis. Clin Cancer Res. 1998;4:799-809.

5. Townsend SE, Allison JP. Tumor rejection after direct costimulation of CD8+ T cells by B7-transfected melanoma cells. Science. 1993;259:368-370.

6. Freedman RS, Tomasovic B, Templin S, et al. Large-scale expansion in interleukin-2 of tumor-infiltrating lymphocytes from patients with ovarian carcinoma for adoptive immunotherapy. J Immunol Methods. 1994;167:145-160.

7. Ioannides CG, Platsoucas CD, Rashed S, et al. Tumor cytolysis by lymphocytes infiltrating ovarian malignant ascites. Cancer Res. 1991;51:4257-4265.

8. Ioannides CG, Freedman RS, Platsoucas CD, et al. Cytotoxic T cell clones isolated from ovarian tumor-infiltrating lymphocytes recognize multiple antigenic epitopes on autologous tumor cells. J Immunol. 1991;146:1700-1707.

9. Kooi S, Freedman RS, Rodriquez-Villanueva J, et al. Cytokine production by T-cell lines derived from tumor-infiltrating lymphocytes from patients with ovarian carcinoma: tumor-specific immune responses and inhibition of antigen-independent cytokine production by ovarian tumor cells. Lymphokine Cytokine Res. 1993;12:429-437.

10. Hefeneider SH, Conlon PJ, Henney CS, et al. In vivo interleukin 2 administration augments the generation of alloreactive cytolytic T lymphocytes and resident natural killer cells. J Immunol. 1983;130:222-227.

11. Henney CS. T-cell-mediated cytolysis: an overview of some current issues. Contemp Top Immunobiol. 1977;7:245-272.

12. Ohta M, Mitomi T, Kimura M, et al. Anomalies in transgenic mice carrying the human interleukin-2 gene. Tokai J Exp Clin Med. 1990;15:307-315.

13. Hendrzak JA, Brunda MJ. Interleukin-12. Biologic activity, therapeutic utility, and role in disease. Lab Invest. 1995;72:619-637.

14. Strieter RM, Kunkel SL, Arenberg DA, et al. Interferon gamma-inducible protein 10 (IP-10), a member of the C-X-C chemokine family, is an inhibitor of angiogenesis. Biochem Biophys Res Commun. 1995;210:51-57.

15. Lenzi R, Kudelka AP, Verschraegen C, et al. Intraperitoneal (IP) bioimmunologic responses in patients with ovarian and gastrointestinal cancers at a low toxicity dosing schedule of IP recombinant interleukin-12 (rhIL-12). Proc Annu Meet Am Assoc Cancer Res. 1999;40:A3778. Abstract.

16. Loercher A, Kavanagh JJ, Verschraegen C, et al. Induction of IL-10 secreting monocytes using peripheral blood mononuclear leukocytes of patients with epithelial ovarian cancer (EOC) cultured in the presence of rGM-CSF and rIL-2. Proc Annu Meet Am Assoc Cancer Res. 1999;40:A3776. Abstract.

17. Wiktor TJ, Macfarlan RI, Reagan KJ, et al. Protection from rabies by a vaccinia virus recombinant containing the rabies virus glycoprotein gene. Proc Natl Acad Sci U S A. 1984;81:7194-7198.

18. Jacobs WR Jr, Tuckman M, Bloom BR. Introduction of foreign DNA into mycobacteria using a shuttle plasmid. Nature. 1987;327:532-535.

19. Aldovini A, Young RA. Humoral and cell-mediated immune responses to live recombinant BCG-HIV vaccines. Nature. 1991;351:479-482.

20. Schmaljohn CS, Hasty SE, Dalrymple JM. Preparation of candidate vaccinia-vectored vaccines for haemmorhagic fever with renal syndrome. Vaccine. 1992;10:10-13.

21. Freedman RS, Lenzi R, Kudelka AP, et al. Intraperitoneal immunotherapy of peritoneal carcinomatosis. Cytokines Cell Mol Ther. 1998;4:121-140.

22. Freedman RS, Edwards CL, Bowen JM, et al. Viral oncolysates in patients with advanced ovarian cancer. Gynecol Oncol. 1988;29:337-347.

23. Freedman RS, Bowen JM, Lotzová E, et al. Virus augmentation as a biologic-modifier approach: experience with intracavitary virus-augmentation therapy. In: Rutledge FN, Freedman RS, Gershenson DM, eds. Gynecologic Cancer: Diagnosis and Treatment Strategies. 1st ed. Austin, Tex: University of Texas Press; 1987.

24. Marshall JL, Hawkins MJ, Tsang KY, et al. Phase I study in cancer patients of a replication-defective avipox recombinant vaccine that expresses human carcinoembryonic antigen. J Clin Oncol. 1999;17:332-337.

25. Ertl HC, Xiang Z. Novel vaccine approaches. J Immunol. 1996;156:3579-3582.

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

27. Freedman RS, Kudelka AP, Verschraegen CF, et al. Therapeutic anticancer vaccine: a randomized double blind dose comparison study of sialyl Tn-KLH with detox-B SE adjuvant for active specific immunotherapy of ovarian cancer (OC). Proc Annu Meet Am Soc Clin Oncol. 1998;17:A1409. Abstract.

28. Tang DC, DeVit M, Johnston SA. Genetic immunization is a simple method for eliciting an immune response. Nature. 1992;356:152-154.

29. Sarzotti M, Dean TA, Remington MP, et al. Induction of cytotoxic T cell responses in newborn mice by DNA immunization. Vaccine. 1997;15:795-797.

30. Xiang Z, Ertl HC. Manipulation of the immune response to a plasmid-encoded viral antigen by coinoculation with plasmids expressing cytokines. Immunity. 1995;2:129-135.

31. Butts C, Zubkoff I, Robbins DS, et al. DNA immunization of infants: potential and limitations. Vaccine. 1998;16:1444-1449.

32. Hurpin C, Rotarioa C, Bisceglia H, et al. The mode of presentation and route of administration are critical for the induction of immune responses to p53 and antitumor immunity. Vaccine. 1998;16:208-215.

33. Nash MA, Lenzi R, Platsoucas CD, et al. RT-PCR quantitation of cytokine responses in vivo from specimens containing small numbers of cells during bioimmunotherapy. J Immunol Methods. 1998;219:169-179.