Cancer Vaccines: Asymptotically Approaching Product Approval

ABSTRACT

The notion that the immune system can prevent the emergence and growth of cancer is a century old. Despite considerable evidence in animal models that "cancer vaccines" induce effective immunological responses against tumors, the development of efficacious vaccines to treat cancer in humans has been much slower and far less promising. In the past decade, however, several vaccine strategies have made considerable clinical advancement, renewing the interest and confidence that a human cancer vaccine will be successfully launched. This article highlights several vaccines that have advanced to later stages of clinical and commercial development. Emphasis is on the development of vaccines using the cancer patient's own autologous antigen-presenting cells (APCs). Mature APCs play a pivotal role in initiating an immune response, especially T-cell mediated immunity, which is critical for the effective killing of tumor cells. Employing APCs in a vaccine to present tumor-associated antigens and stimulate cytotoxic T-cells will enhance the clinical efficacy of the cancer vaccine. As personalized therapeutic products, autologous APC vaccines pose interesting clinical development, regulatory approval, and commercial manufacturing and distribution challenges. Strategies to address some of these challenges are discussed.

It is impossible to cover all of the recent developments in cancer vaccines in this brief review. Therefore, the objective is to present a broad overview of the field by providing examples of a few cancer vaccines currently in commercial development. Emphasis is on the challenges of developing these vaccines.

Immunological Basis for Cancer Vaccines

The immune system can protect against cancer, as well as allow tumors to escape immune destruction.¹ Antibodies and various effector cells, e.g., cytotoxic T-cells (CTLs) and natural killer cells, can recognize and kill tumor cells. Figure 1 depicts the basic mechanisms of the tumor immune response. Most cancer cells express tumor associated antigens (TAAs) that distinguish them as "foreign" tissue to the immune system. Some examples of TAAs targeted by cancer vaccines include: MART-1, MAGE-3, NY-ESO-1, prostate specific antigen (PSA), and prostatic acid phosphatase (PAP). Soluble and membrane-bound TAAs are captured and processed by antigen-presenting cells (APCs) which carry them to lymph nodes that drain the tumor site. Immunogenic fragments (epitopes) of processed TAAs are associated with the major histocompatability complex (MHC) molecules on the surface of the APCs. In concert with various cyotokines, the TAA-loaded APCs present the antigens to T-cells (thymus derived lymphocytes) and/or B-cells (bone marrow derived lymphocytes), which activates the T and B cells to proliferate and differentiate into immune effector cells. B-cells differentiate into plasmacytes that secrete soluble, antigen-specific antibodies that mediate killing of cancer cells by binding to the membrane-bound tumor antigens to activate the compliment cascade, or attracting antibody-dependent cytotoxic cells to the tumor. T-cells can differentiate into TAA-specific CTLs that migrate from the lymph nodes to the site of the tumor, where they kill the tumor through the release of cytotoxic enzymes such as perforin and granzyme B. Other cells that participate in a tumor immune response include helper T-cells that release cytokines to activate CTLs and B-cells, and regulatory T-cells that can allow tumor immune escape by inhibiting an effective tumor immune response. Cancer vaccines are intended to introduce TAAs to the patient in a way that stimulates a potent tumor immune response.

Figure 1. The tumor immune response

Cancer Vaccine Platforms

Some approved cancer therapies employ immunological mediates, but are not considered vaccines. Interleukin-2 (IL-2, Proleukin for treating melanoma and renal cell carcinoma),² alpha-interferon (IFN-α, Intron A to treat various cancers)³ and monoclonal antibodies (e.g., Trastuzumib, which binds the HER2 epidermal growth factor receptor on breast cancer)⁴ are familiar cancer therapy agents. For this review, however, the focus is on cancer vaccines, which as defined above, introduce TAAs to the patient to elicit an effective tumor immune response. Two categories of cancer vaccines that differ in their technical complexity will be discussed.

Tumor Antigen–Based Vaccines

Tumor antigen–based vaccines can be viewed as traditional vaccines that are approved for prevention of infectious diseases. The vaccine consists of an antigen formulated with or without an adjuvant or immunostimulator to enhance immunogenicity. Recombinant proteins, peptides, soluble tumor lysates, and killed tumor cells are used as TAAs in these cancer vaccines. The simplest of these vaccines consists of full length protein or peptide TAAs in a suitable pharmaceutical vehicle for parenteral delivery. Examples of protein- or peptide-based cancer vaccines in clinical development include:

• IDM Pharma's EP-2101 vaccine, which consists of nine synthetic peptides based on epitopes from four TAAs for treatment of non-small cell lung cancer. Promising interim results from a 68-patient Phase 2 clinical trial were recently reported.⁵

• PROSTVAC-VF, the Therion Biologies vaccine for treatment of prostate cancer.⁶ The vaccine consists of two genetically engineered viruses (vaccinia and fowlpox given as prime and boost vaccinations, respectively) that express PSA and three co-stimulatory molecules, B7.1, ICAM-1, and LFA-3 (TRICOM).

• The Favrille vaccine (Favld, for treatment of B-cell non-Hodgkin's lymphoma) uses genetic engineering to produce a recombinant protein of the patient's B-cell idiotype, the TAAs specific to that cancer.⁷ The idiotype antigen is conjugated with keyhole limpet hemocyanin (KLH), an immunostimulatory molecule that is then formulated with granulocyte macrophage colony stimulating factor (GM-CSF), which is a potent stimulator of APCs.

Autologous and allogeneic (patient unrelated) tumor cells can be used as cancer vaccines. For these vaccines, either a crude, soluble cell lysate or whole, killed tumor cells are formulated for injection. Tumor-cell vaccines are often a mixture of TAAs and nonantigenic cellular components. An example of this type of vaccine is Cell Genesys's GVAX for prostate cancer. This vaccine comprises two prostate cancer cell lines, LNCaP and PC-3, which, in addition to expressing several TAAs, were transduced to express GM-CSF. In a Phase 1–2 clinical trial, 16/21 (76%) hormone-naive prostate cancer patients had a significant decrease in PSA velocity and increased TAA antibody titers at 12 weeks following eight weekly intradermal injections of the vaccine.⁸ Phase 3 clinical trials with this vaccine are under way. The GVAX technology for autologous and allogeneic tumor cell vaccines is being applied to a variety of cancers.⁹

Cell–Based Vaccines

Cell-based vaccines use ex vivo–prepared TAA-loaded APCs as the vaccine.¹⁰ Table 1 provides an overview of the array of technology platforms. Briefly, the patient's APCs are isolated from peripheral blood cells and placed in cell culture for loading with TAAs. For many of these vaccines, the precursor monocytes or CD34+ cells are cultured with cytokines to obtain dendritic cells (DC), which are very potent APCs.¹² Examples of these types of vaccines include:

Table 1. Summary of dendritic-cell vaccine platforms

• Provenge, Dendreon's vaccine for treating prostate cancer. This vaccine is prepared with the patient's monocytes enriched by elutriation from leukapheresis blood cells. The monocytes are placed in cell culture without exogenous cytokines and loaded with the tumor antigen, which is a fusion protein of full-length PAP and GM-CSF. After 40 hours in culture, the antigen- loaded APCs are washed and formulated for infusion into the patient.¹³ In hormone-refractory prostate cancer patients treated with Provenge (N = 82) the median survival was 25.9 months compared to 21.4 months in the placebo (APCs without antigen) group (N = 45).¹⁴ Dendreon received an FDA approvable letter on May 8, 2007, with licensure pending Phase 3 clinical trial results to support the efficacy claim.¹⁵

• Argos Therapeutics is testing a DC vaccine for the treatment of renal cell carcinoma.¹⁶ The vaccine is monocytes grown with GM-CSF and IL-4, transfected with autologous tumor RNA and matured with IL-1β, IL-6, TNF-α, and PGE₂.¹⁷

• UVIDEM is a DC vaccine for treating melanoma. It is prepared by culturing the patient's monocytes with IL-13 and loading them with lysates from allogeneic melanoma cell lines.¹⁸

• ODC Therapy has licensed DC vaccine technology from the Baylor Research Institute for commercial development. The vaccines are prepared as either CD34+ cells with GM-CSF, TNF-α, Flt-3L and loaded with MART-1, gp-100, tyrosinase and MAGE-3 peptides,¹⁹ or monocytes with GM-CSF, IL-4, CD40L, TNF-α and loaded with killed allogeneic melanoma cells.²⁰ Both vaccines elicited anti-tumor immune responses and prolonged survival in Stage IV melanoma patients.

Commercial Challenges for Cancer Vaccines

Although each of the cancer vaccines mentioned above has unique development and cGMP compliance challenges, they share some common manufacturing and quality control release issues. Figure 2 outlines a general scheme for the manufacture of DC vaccines. This scheme has two distinct paths. One path, which also applies to antigen-based vaccines, depicts the steps involved in manufacture of the TAAs. The other path depicts the steps to manufacture the TAA-loaded DC. Criteria for the manufacture and release of recombinant proteins and synthetic peptides are not discussed here, since they have been established for licensed biological products. It should be noted that regulatory agencies are still considering requirements for the manufacture and release of cell-based cancer vaccines.

Figure 2. General production scheme for manufacture of dendritic-cell cancer vaccines

Tumor Antigen Manufacture and Release

The first step in the TAA preparation process is establishing control over the growth and stability of the allogeneic human tumor cell lines or autologous patient tumor cells. Although numerous allogeneic human tumor cell lines exist, the potential drawbacks to their use as sources of TAAs include: 1) a potential lack of traceability to the original tumor isolate, 2) the possible exposure of the cells to animal products, e.g., serum, used in cell culture, and 3) the difficulty in establishing cGMP manufacturing control and quality compliance. Although cGMP-compliant master and working cell banks can be established for allogeneic tumor cell lines, this is less likely to occur when autologous tumor cells are the source of TAAs, thus presenting a different set of challenges, including: 1) whether enough tumor cells can be harvested from the patient, 2) whether the cells can be established and expanded in cell culture to the numbers needed to prepare the vaccine, 3) whether the cells express the desired TAAs in the amounts needed to prepare a vaccine to meet the vaccination schedule for the patient, and 4) how a cGMP process will be developed that accommodates the manufacture of a consistent vaccine from different, noncharacterized tumor explants.

Production of the tumor antigens is the second step in the process. Manufacturing considerations at this stage of the process include whether the cells: 1) express the TAA of interest, 2) yield enough TAA to manufacture vaccine, 3) can be used to make a desired stable transformed line, and 4) can be grown under acceptable cGMP cell culture conditions. The cell culture growth conditions must be set in regard to the type of culture vessel to be used (e.g., flasks, "cell factories," closed bags, or roller bottles), whether the cells can be adapted to serum-free media or whether traceable animal products are required in the growth media, and the qualification of the manufacturing site for expanding the tumor cell lines and preparing the TAAs.

Manufacturing facility qualification is less of an issue for allogeneic tumor cell lines than for autologous tumor cells, since allogeneic tumor cell banks can be established and a single cGMP manufacturing site can be qualified for production of TAAs to meet commercial demand. In contrast, the use of autologous tumor cells presents several unique problems in maintaining manufacturing control and product consistency. Since the tumor cells are obtained from individual patients, the first decision is whether to prepare the TAAs at the collection site under a qualified procedure or to ship the cells to a central location where the TAAs can be prepared in a validated cGMP manufacturing facility. Regardless of where the vaccine is manufactured, establishing product consistency between TAA products will be difficult when using unique autologous tumor cell isolates.

The final step in the process is QC release of the formulated and filled TAA vaccine product for patient use or drug substance for loading into DC. The following issues should be considered at this step. First, cells maintained in culture for extended periods can become contaminated with mycoplasma, thus, precautions and testing are required to ensure the TAAs are free of mycoplasma. Second, when killed tumor cells are used as the vaccine, the process should be validated to ensure that the tumor cells are not capable of progressive growth in vivo. Third, a decision is needed as to whether dosing will be based on TAA amount or cell number. Finally, TAA lysate and killed tumor cells can be filled into vials and frozen for storage prior to injection, assuming that stability-indicating assays are in place and that the desired stability can be established for the product.

DC Vaccines

The initial steps in the manufacture of a DC vaccine involve enrichment of the precursor cells from the patient's apheresis blood product. Once isolated, the DC precursor cells are placed in cell culture with the appropriate mixture of cytokines and loaded with TAAs. Upon completion of the cell culture step, the DC are harvested, formulated, and filled for injection into the patient. The vaccine can be given to the patient as either freshly prepared DC or from a batch of frozen DC.

Since the DC are derived from the patient, processed ex vivo, and formulated for injection, the same issues outlined above for the use of autologous tumor cells for the manufacture of TAAs apply here to the preparation of autologous DC vaccines. These include: how to qualify a cGMP facility(ies) to manufacture vaccines with cells collected at various locations; how to validate manufacturing controls for a vaccine product based on unique blood cell populations; and how to establish vaccine consistency for a product derived from individual patients. Whether the DC vaccine needs to be made fresh or can be frozen requires considerable characterization of the product and correlation with clinical efficacy data.

Quality Control and Release Testing

Cancer vaccines, whether TAA- or DC- based, must be released to the same criteria as those required for other pharmaceuticals, i.e., sterility, identity, purity, and potency. Sterility testing must establish that these products are negative for gram and Hoechst stains, and there is no outgrowth of mycoplasma and bacteria at 14 days and fungus at 30 days. Since expediency is key to the treatment of cancer patients, initial vaccine dosing may begin before sterility testing is completed. In those instances, the vaccine may be released with a minimum amount of sterility testing completed, usually a gram stain, Hoechst stain, and PCR testing for mycoplasma. The patient must be monitored for symptoms of infectious diseases and appropriate treatment initiated if necessary. Therefore, validating a sterile process early in the development of these products is advisable.

Validating various QC tests for release of cancer vaccines may be difficult. Cell number and viability may be used to determine the dose of killed tumor cell or DC vaccines. Manual cell count and viability methods can be highly variable and automated methods can be difficult to establish. Identity and purity tests will likely be based on analysis of TAAs and cell surface markers expressed by tumor cell lines or TAA-loaded DC. For TAAs proteins and peptides, validated physical or chemical methods are known, e.g., HPLC or ELISA. Microscopy and flow cytometric methods can be validated to quantify TAA and cell marker expression on tumor cell lines and DC. Cell surface markers, such as CD1, CD14, CD80, CD83 and HLA-DR, are frequently used to identify and quantify DC in a vaccine. Antigen loading can be analyzed if reagents are available that recognize TAAs within the various antigen-processing cellular compartments or as epitopes associated with MHC on the surface of DC. Determining the correct proportion of specific DC subpopulations required for an effective vaccine must be established in clinical studies and is likely to result in a broad specification for DC vaccines. Potency, defined in 21 CFR 600.3(s) as the specific ability or capacity of the product to effect a given result, is a more difficult parameter to address for the release of cancer vaccines. Since the mechanism of action of a vaccine is to elicit an effective immune response, it is a daunting, if not impossible, task to develop and validate a practical in vitro immunological potency assay for QC release. Although mixed lymphocyte response assays can measure proliferation of allogeneic or autologous T-cells stimulated by APCs in a vaccine, validation of these complex biological assays is unlikely. Thus, most sponsors may seek to release product based on potency assays using surrogate markers. For example, increased expression of CD54 by TAA-loaded APCs is the potency assay used by Dendreon to release the Provenge vaccine. The reliability of a surrogate marker as an indication of vaccine potency will depend on a strong correlation with nonclinical analyses and clinical data gathered during product development.

Future Considerations

The commercial manufacturing and regulatory approval pathways for cancer vaccines will become clearer as several of these products move closer to licensure. However, for "personalized" vaccines, extensive clinical data may be required to demonstrate that the manufacturing process is well controlled and produces a consistent product. Although cancer vaccines under development have proven safe and well tolerated, establishing clinical efficacy remains a challenge. For a cancer vaccine, it is imperative to establish a correlation between immunological endpoints and clinical responses, and to demonstrate a survival advantage over the standards of care in randomized Phase III clinical trials. Successful commercial development of cancer vaccines may in fact require new approaches to the design of clinical trials,²¹ as well as the manufacture and release of pharmaceutical products.

Acknowledgements

Thanks to Dr. Karen Auditore-Hargreaves, CEO of ODC Therapy, Inc., for her helpful comments on the preparation of this article. My sincerest apologies to all those companies and investigators whose work in this field I was not able to mention due to space constraints. Your efforts are recognized as important contribution to the eventual successful development of effective cancer vaccines.

Lee K. Roberts, PhD, is the vice president of operations at ODC Therapy, Inc., Dallas, TX, 214.370.6181, lee.roberts@odctherapy.com.

References

1. Dunn GP, Bruce AT, Ikeda H, Old LJ, Schreiber RD. Cancer immunoediting: from immunosurveillance to tumor escape. Nature Immunol. 2002;3(11):991–8.

2. Proleukin is marketed by Novartis Pharmaceuticals Corporation and information about this product can be found at: www.proleukin.com.

3. Schering-Plough webpage; www.schering-plough.com. See cancer therapies under Products & Care for description of Intron A.

4. Reichert JM, Valge-Archer VE. Development trends for monoclonal antibody cancer therapeutics. Nature Reviews/Drug Discovery. 2007;6:349–56.

5. IDM Pharma webpage; www.idm-biotech.com. For a EP2101 product description see Products, Science & Technologies/Product Candidates/Products to Prevent Recurrence, and for interim clinical results open Financial Information & News Desk/News Release to see 06/04/07 IDM Pharma presents interim Phase II data from EP2101 lung cancer vaccine clinical trial.

6. DiPaola RS, Plante M, Kaufman H, Petrylak DP, Israeli R, Lattime E, Manson K, Schuetz T. A phase I trial of Pox PSA vaccines (PROSTVAC-VF) with B7-1, ICAM-1, and LFA-3 co-stimulatory molecules (TRICOM) in patients with prostate cancer. J Translational Med. 2006;4:1–5.

7. Favrille webpage; www.favrille.com for product information under the Our Technology tab.

8. Simons JW, Carducci MA, Mikhak B, Lim M, Biedrzycki B, Borellini F, Clift SM, Hege KM, Ando DG, Piantadosi S, Mulligan R, Nelson WG. Phase I/II trail of an allogeneic cellular immunotherapy in hormone-naive prostate cancer. Clin Cancer Res. 2006;12(11):3394–3401.

9. Hege KM, Jooss K, Pardoll D. GM-CSF gene-modified cancer cell immunotherapies: Of mice and men. Intl Rev of Immunol. 2006;25:321-52.

10. Nestle FO, Farkas A, Conrad C. Dendritic-cell-based vaccination against cancer. Current Opinion Immunol. 2005;17:163–9.

11. Ridgway D. The first 1000 dendritic cell vaccinees. Cancer Invest. 2003;21:873–86.

12. Banchereau J, Palucka AK. Dendritic cells as therapeutic vaccines against cancer. Nature Reviews/Immunol. 2005; 5:296–306.

13. Small EJ, Fratesi P, Reese DM, Strang G, Laus R, Peshwa MV, Valone FH. J Clin Oncol. 2000;18(23):3894–3903.

14. Small EJ, Schellhammer PF, Higano CS, Redfern CH, Nemunaitis JJ, Valone FH, Verjee SS, Jones LA, Hershberg RM. Placebo-controlled phase III trail of immunologic therapy with Sepuleucel-T (APC8015) in patients with metastatic asymptomatic hormone refractory prostate cancer. J Clin Oncol. 2006;24(19):3089–94.

15. Dendreon Corporation webpage; www.dendreon.com see May 9,2007, Dendreon receives complete response letter form FDA for Provenge biologics license application, under Newsroom/Press Releases.

16. Su Z, Dannull J, Heiser A, Yancey D, Pruitt S, Madden J, Coleman D, Niedzwiecki D, Gilboa E, Vieweg J. Immunological and clinical responses in metastatic renal cancer patients vaccinated with tumor RNA-transfected dendritic cells. Cancer Res. 2003;63:2127–33.

17. Harris J, Monesmith T, Ubben A, Norris M, Freedman JH, Tcherepanova I. An improved RNA amplification procedure results in increased yield of autologous RNA transfected dendritic cell-based vaccine. Biochim Biophys ACTA. 2005;1724:127–36.

18. IDM Pharma webpage; www.idm-biotech.com for the UVIDEM product description see Products, Science & Technologies/Product Candidates/Products to Prevent Recurrence, and for interim clinical results open Financial Information & News Desk/News Release to see 06/05/07 IDM Pharma presents preliminary results form Phase II UVIDEM melanoma vaccine clinical trial.

19. Banchereau J, Palucka AK, Dhodapkar M, Burkeholder S, Taquet N, Rolland A, Taquet S, Coquery S, Wittkowski KM, Bhardwaj N, Pineiro L, Steinman R, Fay J. Immune and clinical responses in patients with metastatic melanoma to CD34+ progenitor-derived dendritic cell vaccine. Cancer Res. 2001;61:6451–58.

20. Palucka AK, Ueno H, Connolly J, Kerneis-Norvell F, Blanck JP, Johnston DA, Fay J, Banchereau J. Dendritic cell loaded with killed allogeneic melanoma cells can induce objective clinical responses and MART-1 specific CD8+ T-cell immunity. J Immunother. 2006;29:545–7.

21. Hoo A, Parmiani G, Hege K, Sznol M, Loibner H, Eggermont A, Urba W, Blumenstein B, Sacks N, Keilholz U, Nichol G. A clinical development paradigm for cancer vaccines and related biologics. J Immunother. 2007;30(1):1–15.