Allogeneic tumor cell vaccines

Abstract

The high mortality rate associated with cancer and its resistance to conventional treatments such as radiation and chemotherapy has led to the investigation of a variety of anti-cancer immunotherapies. The development of novel immunotherapies has been bolstered by the discovery of tumor-associated antigens (TAAs), through gene sequencing and proteomics. One such immunotherapy employs established allogeneic human cancer cell lines to induce antitumor immunity in patients through TAA presentation. Allogeneic cancer immunotherapies are desirable in a clinical setting due to their ease of production and availability. This review aims to summarize clinical trials of allogeneic tumor immunotherapies in various cancer types. To date, clinical trials have shown limited success due potentially to extensive degrees of inter- and intra-tumoral heterogeneity found among cancer patients. However, these clinical results provide guidance for the rational design and creation of more effective allogeneic tumor immunotherapies for use as monotherapies or in combination with other therapies.

Keywords: :

• immunotherapy
• cancer
• allogeneic
• vaccine
• adjuvant
• clinical trials

Introduction

Cancer is characterized by the uncontrolled proliferation of the body’s cells.^1,2 Initiated to a great extent by mutations in the genome affecting cell regulatory function, potential cancers arise throughout the body constantly.³ However, natural processes such as apoptosis and immune surveillance eradicate these cancers before the development of tumors.^4-10 Despite the protection it confers, immune surveillance directs the selection of malignancies. This process, termed immunoediting, explains the mechanism by which cancers arise in immune competent individuals. Briefly, malignant cells recognized by the immune system are eliminated, while unrecognized malignancies survive. The repeated selection of these immunologically silent, cancerous cells leads to the eventual formation of a tumor mass. This tumor microenvironment is immunosuppressive due to the secretion of inhibitory cytokines, expression of inhibitory membrane bound ligands and recruitment of regulatory cell populations that deactivate tumor-specific immune cells.^5,11,12 These factors create an environment which promotes cancer growth, unhindered by the body’s natural defenses.⁵

Although current cancer therapies such as chemotherapeutic agents and ionizing radiation decrease the percentage of relapse in some types of cancers, the clones that emerge after treatment may be more resistant to the same treatments. Further, these therapies are also associated with either a lack of specificity in the case of chemotherapy or a loss of hematopoietic potential following radiation therapy.¹³ While these problems remain unsolved, activation of the immune system against cancer remains as a viable therapeutic option.^14-17 As a result, a number of immunotherapeutic treatments have thus been explored to augment a patient’s natural anti-tumor immune response and overcome tumor-induced immunosuppression. These therapies, varying greatly in composition, aim to induce specific anti-tumor immunity.

Allogeneic vs. Autologous Therapeutic Cancer Vaccines

Therapeutic vaccination strategies for cancer can be categorized broadly by antigen source. From an immunological standpoint, successful cancer therapeutics stimulate the immune system against a broad range of tumor-associated antigens (TAAs) while conferring lasting immunity. Other considerations include vaccine efficacy over multiple doses, vaccine kinetics, and specificity. The source of biological material for a vaccine can modulate these properties and influence the vaccine’s overall efficacy.

Autologous cancer vaccination strategies use material derived from a patient’s tumor to create personalized treatments. These treatments vary over a broad spectrum from immunization with patient-specific purified cancer antigens, to the induction of anti-tumor immunity with irradiated, patient-derived, cancer cells. Cell-based therapies can be further specialized by transfection with additional immune stimulatory molecules.¹⁸ These autologous immunotherapies contain a wide array of proteins and peptides specific to the individual’s tumor antigens presented by self major histocompatibility complex (MHC) to stimulate the immune system directly.¹⁹ Moreover, recent studies have demonstrated that tumors are made up of a highly heterogeneous population of cancer cells.^20,21 Therefore, it can be hypothesized that autologous cell lines may be insufficient in the elicitation of a broad immune response and an entire primary tumor mass may be required for effective vaccination.

Allogeneic vaccines are largely similar to autologous vaccines with the exception that material is sourced from another member of the same species. Commonly used allogeneic materials include use of established laboratory-grown cancer cell lines known to express TAAs of a specific tumor type. This allows these therapeutics to be mass-produced, stored, and modified prior to use.²² Although many allogeneic cell based vaccines are from cancer cell lines of the same species and type (breast, prostate, lung, etc.), they may differ from autologous vaccines in that they do not contain “patient-specific” tumor antigens.²³ Despite this disadvantage, preclinical studies in prostate cancer²⁴ and melanoma²⁵ models have demonstrated an added benefit of allogeneic sourced vaccines; presentation of TAAs in an allogeneic setting increases immunogenicity. Although, the supporting data are largely preclinical,²⁴ the hypothesis that allogeneic cells provide an additional danger signal is consistent with increased graft vs. tumor responses observed in patients who receive mismatched minor histocompatibility antigen bone marrow transplants.²⁶ Moreover, other desirable properties including availability, low production cost, and the lack of invasive procedures make allogeneic immunotherapy a salient modality.²⁷

Allogeneic Tumor-Based Vaccination Strategies

Allogeneic or autologous tumor vaccines can be prepared and administered to patients using irradiated whole cells or cell lysates. The use of these vaccine modalities allow for a reproducible, safe vaccine product in which injected tumor cells cannot replicate.²⁸ Additionally, irradiated cells naturally express and present numerous TAAs, alleviating the need to identify and purify TAAs, and thus promoting the initiation of an anti-tumor specific immune response.^19,29 It has also been shown that immune recognition of tumor cells by CD8⁺T cells and antigen-presenting cells (APCs) is enhanced following irradiation.^30,31 Furthermore, whole cell vaccines can be modified to increase immunogenicity by transfection of immune stimulatory molecules ex vivo.^32,33 For example, the GVAX vaccine platform involves the use of irradiated allogeneic tumor cell lines, modified to secrete the cytokine granulocyte-macrophage colony-stimulating factor (GM-CSF) to augment the immunogenicity of the vaccine.^32,34

Despite these advantages, drawbacks of using irradiated whole tumor cells as a vaccine also exist. For example, upon irradiation of cells, phosphatidylserine (PS), an immunosuppressive phospholipid usually found on the inner leaflet of the plasma membrane, is transferred to the outer leaflet. Expression of PS on the cell surface of irradiated cells has been shown to induce secretion of immunosuppressive factors by dendritic cells (DC) and inhibit maturation, promoting an immunosuppressive tumor environment. Furthermore, irradiated cells may retain the ability to secrete immunosuppressive factors like the original tumor cells.^35,36 Thus irradiation induced immune suppression, partially negates the immunogenic effect of irradiated whole cells.³⁷ To overcome the immunosuppressive nature of irradiated whole tumor cell vaccines, Huang et al. generated anti-PS antibody:IL-2 immunocytokine complexes. The role of PS in immunosuppression and the ability of this biomolecule to block PS were confirmed by increased vaccine-specific immunogenicity in mice vaccinated with irradiated cells.^37,38

Tumor cell lysates have also been used as vaccines to stimulate anti-tumor immune responses.³⁹ Although these vaccines allow for the presentation of multiple TAAs, they exhibit poor immunogenicity and some have been shown to be completely ineffective. The inability of tumor cell lysates to stimulate a sustained immune response may be due in part to the presence of immune suppressive molecules within the lysate.⁴⁰ To offset these deficiencies, many attempts have been made to enhance immunogenicity through the administration of adjuvant along with the vaccines. For example, Dong et al., showed that tumor cell lysates from Lewis lung cancer cells mixed with mycobacterial heat shock protein (HSP)65 were able to inhibit tumor growth and prolong the survival of lung cancer bearing mice by the activation of tumor specific cytotoxic T lymphocytes (CTLs) as well as innate immune cells.⁴¹

Another whole cell lysate vaccination strategy utilizes DCs and their ability to efficiently capture and present antigen to T cells. This method, known as DC pulsing, allows for the activation of tumor specific CD4⁺ and CD8⁺ T cells through the MHC:peptide TCR interaction.^42,43 DCs serve as an ideal link to the adaptive immune system due to high expression of MHC class I and II and costimulatory molecules and the subsequent initiation of an immune response. Moreover, DCs are also able to cross present exogenously loaded antigen and induce CD8⁺ T cell responses. These characteristics make DCs ideal for ex vivo loading of tumor cell lysates as evidenced by clinical studies in which administration of DC pulsed vaccines generates a large repertoire of tumor-specific immune responses^44,45

Clinical Trials with Allogeneic Cancer Vaccines

The following section provides a comprehensive look at results from clinical trials for various cancers in which disease stabilization is often achieved in conjunction with vaccine-induced immune responses. The results of these studies are summarized in Table 1.

Table 1. Overview of findings from allogeneic clinical trials for cancer immunotherapies

Cancer origin	Type of therapy	Phase	Total patients	Median overall survival	Reference
Breast	Allogeneic GM-CSF-secreting whole cell breast cancer vaccine	I	60	—	Emens, Armstrong 2004^108
Breast	Allogeneic costimulatory gene (CD80)-modified, HLA-A2-matched, breast cancer cell line	I	30	—	Dols, Smith et al. 2003^52
Breast	Allogeneic GM-CSF-secreting breast cancer cell lines (SKBR3 and T47D) alone or with CY and DOX	I	28	—	Emens, Asquith et al. 2009^109
Chronic myeloid leukemia	CML cell line modified to secrete GM-CSF administered with imatinib mesylate	I	19	—	Smith, Kasamon et al. 2010^61
Lung	Autologous DCs, generated by culturing CD14+ cells with GM-CSF and IL-4, were cultured with apoptotic bodies from 1650-TC expressing adenocarcinoma cell line, and then administered i.d.	I	14	—	Yannelli et al. 2005,^110 Hirschowitz et al. 2007^111
Lung	K562 cells genetically modified to secrete GM-CSF combined with autologous long tumor cells	I/II	86	5.4 mo	Nemunaitis, Jahan et al. 2006^112
Lung (NSCLC)	Administration of Belagenpumatucel-L (cocktail of 4 irradiated allogeneic NSCLC cell lines transfected with TGF-β2 antisense transgene)	II	75	14.5 mo	Nemunaitis et al. 2006^56
Lung (NSCLC)	Adminstration of irradiated whole cell (AD100) allogeneic vaccine transfected to express B7.1 along with either HLA-A1 or HLA-A2	I	19	18 mo	Raez, Cassileth et al. 2004^55
Lung (NSCLC)	Allogeneic tumor cells secreting endoplasmic reticulum-chaperone gp96-Ig-peptide complexes	I	19	16.5 mo	Raez, Walker et al. 2013^59
Melanoma	Administration of 3 irradiated allogeneic Cell lines with BCG	III	670	Discontinued No clinical benefit	—
Melanoma	Administration of Mel-D and Mel-S cell lysates (Melacine) with DETOX	I	17	—	Mitchell, Kanmitchell et al. 1988^64
Melanoma	Administration of Melacine with CY and IFN-α i.v. after 4 doses of Melacine	II, III	47	12.5 mo	Vaishampayan, Abrams et al. 2002^65
Melanoma	Melacine administration—investigate impact of class I antigen expression on relapse-free survival after adjuvant therapy with vaccine	III	689	5 y (relapse-free survival)	Sosman, Unger et al. 2002^66
Melanoma	Ex vivo loading of autologous DCs with antigens from apoptotic/necrotic allogeneic melanoma cells and subsequent adoptive transfer	I	16	—	Von Euw, Barrio et al. 2008^113
Melanoma	Administration of Cyp followed by VACCIMEL (mixture of 3 allogeneic cell lines)	II	30	—	Mordoh et al. 1997^114
Melanoma	Administration of VACCIMEL with rhGM-CSF	I	20	—	Barrio et al., 2006^115
Melanoma	Administration of autologous monocyte derived DCs loaded ex vivo with killed allogeneic Colo829 melanoma cells and activated with GM-CSF, IL-4, TNF-α and CD40 ligand	I	21	22.5 mo	Palucka, Ueno et al. 2006^69
Pancreatic cancer	Administration of allogeneic GM-CSF-secreting tumor vaccine eight weeks post pancreaticoduodenectomy	I	14	—	Jaffee et al. 2001^90
Prostate	Administration of 3 irradiated prostate cancer cell lines with BCG	IIb	44	—	—
Prostate	First administration of an allogeneic prostate cancer cell lines (PC3 and LNCap) modified to secrete GM-CSF (based on GVAX platform)	I/II	21	—	Simons, Carducci 2006^116
Prostate	Administration of an allogeneic prostate cancer cell lines (PC3 and LNCap) modified to secrete GM-CSF	I/II	34	34.9 mo (high dose); 24.0 mo (low dose)	Small, Sacks, Nemunaitis 2007^73
Prostate	Administration of GVAX plus Ipilimumab (fully human IgG CTLA-4 blocking Ab)	I	12	29.2 mo	Van den Eertwegh, Versluis, van den Berg 2012^117
Prostate	Administration of 2 allogeneic prostate-carcinoma cell lines that are genetically engineered to secrete GM-CSF	I/II	80	35.0 mo (high-dose); 20.0 mo (mid-dose), 23.1 mo (low-dose)	Higano, Corman, Smith et al. 2008^77
Prostate	Administration of LNCaP modified using retroviral vector to secrete IL-2 and IFN-γ	I	6	—	Brill, Kuebler, Pohla et al. 2007^118
Prostate	Administration of LNCaP modified using retroviral vector to secrete IL-2 and IFN-γ	I/II	30	32 mo	Brill, Kuebler, Pohla, Buchner 2009^119
Prostate	Administration of whole cell vaccine consisting of a mixture of 3 prostate cancer cell lines along with Mycobacterium vaccae (SRL172)	I/II	60	—	Eaton, Perry et al. 2002^71
Colon	Administration of 3 melanoma cell lines with BCG	I	27	12.9 mo	Habal, Gupta et al. 2001^93
Renal	Administration of two renal cancer cell lines genetically modified to express IL-7 and B7-1	I	10	40 mo	Westermen et al. 2011^94

A summary of clinical trials for allogeneic cancer immunotherapies lists the type of cancer, type of therapy, clinical trial phase, patients enlisted, median survival (if listed), and reference.

Breast Cancer

Due to the lack of effective treatment options for patients with late stage or metastatic breast cancer⁴⁶ and the difficulty of culturing primary cancer cells, allogeneic whole cell vaccines are currently being investigated as a potential adjuvant therapy to chemotherapy. Vaccines consisting of genetically modified tumor cells secreting various cytokines, such as GM-CSF, have previously been shown to be safe and induce an immune response in phase I and phase II clinical trials.⁴⁷ However, these responses were not sufficient to overcome the immune suppression induced by established tumors. In order to be effective as a monotherapy, cancer vaccines must be capable of inducing potent and sustainable tumor-specific immune responses that will reduce overall tumor burden. This provides a growing precedent for effectively combining immunotherapies with cytotoxic agents.¹⁴

In a phase I clinical trial enrolling 28 patients with metastatic breast cancer, the safety and clinical efficacy of an allogeneic GM-CSF-secreting breast cancer vaccine was investigated. The vaccine, formulated from two HER2/neu positive mammary adenocarcinoma breast cancer cell lines (SKBR3 and T47D)⁴⁸ were administered either alone or in sequence with common chemotherapeutic agents cyclophosphamide (CY) and doxorubicin (DOX).⁴⁷ The goals of the study were to determine the safety of this combinatorial therapy and the optimal chemotherapy dose that would induce enhanced immunity against the breast cancer TAA HER2.⁴⁹

This study showed that up to 4 doses were well tolerated by patients, while accompanied by only modest levels of toxicity that was not exacerbated with the addition of chemotherapy. Additionally, the vaccine either alone or in sequence with low-dose chemotherapy (200 mg/m² CY) could induce HER2-specific T-cell mediated immunity. HER2-specific humoral immunity was enhanced in the serum of patients receiving the vaccine along with 200 mg/m² CY or 35 mg/m² DOX.⁴⁷

These results suggest that low dose chemotherapy can be used to break tolerance, while sustaining an antigen-specific immune response. Moreover, these results are consistent with a working hypothesis that low dose chemotherapy can delay the vaccine specific immune response witnessed after repeated vaccination with allogeneic cells. When taken with preclinical data that DOX¹⁴ and CY⁵⁰ augment anti-tumor immune responses while depleting Tregs,⁵¹ a combinatorial approach of chemotherapy and immunotherapy should be considered strongly as a multivalent therapeutic modality.

In an another phase I study, the costimulatory molecule B7-1 (CD80) was transfected into HLA-A2⁺ matched allogeneic MDA-MB-231 breast cancer cell line and used as a vaccine to treat stage IV breast cancer patients.⁵² Although vaccinated patients showed an increased tumor-specific immune response there was no tumor regression. However, the administration of a combination of allogeneic breast cancer cells (MCF-7), autologous breast cancer cells, TAA and low doses of systemic GM-CSF and IL-2 showed both a significant increase in antigen-specific lymphocyte response and an improved clinical response.⁵³ A separate study also reported that a similar multi-antigen combination vaccination consisting of TAAs, allogeneic, and autologous cells with systemic administration of biological adjuvants improved 10-y survival of breast cancer patients significantly in patients with depressed lymphocyte immunity at the start of the treatment (59% for unvaccinated group compared with 89% vaccinated group).⁵⁴ These observations suggest that a multi-antigen vaccination approach in combination with biological adjuvants is another promising therapeutic approach for breast cancer.

Lung Cancer

Counteracting the immunosuppressive environment created by primary tumors, such as those formed in non-small cell lung cancers (NSCLC), presents an opportunity for immune mediated disease stabilization. Recent developments including the identification of lung cancer TAAs (Table 2), immunosuppressive factors, and the establishment of multiple tumor cell lines have contributed to the growing number of available reagents.⁵⁵ The immunotherapeutic Belagenpumatucel-L (Lucanix), which is currently undergoing randomized phase III, utilizes multiple allogeneic tumor cell lines. Results from the prior phase II clinical trial conducted by Nemunaitis et al. demonstrated that subgroup analysis of individuals with anti-vaccine immune responses had an increased overall survival (OS).⁵⁶ Enrolling 75 patients with both early and late stage (stage II, IIIA, IIIB, and IV) NSCLC, patients were vaccinated with Lucanix, a cocktail of two adenocarcinomas, one squamous carcinoma, and one large-cell carcinoma cell lines (NCI-H-460, NCI-H-520, SK-LU-1, and Rh-2) transfected with a TGF- β2 antisense transgene to knockdown TGF-β2. Because TGF- β2 is a cytokine known to mediate immunosuppression high levels and correlate with poor prognosis, knock down of TGF- β2 could activate an anti-tumor immune response. This trial resulted in a one-year survival rate of 68% and a two-year survival rate of 52% for patients receiving the higher vaccine dose and 39% and 20%, respectively, for patients receiving a lower dose. Serum cytokine analysis determined significantly increased levels of IFN-γ, IL-6, and IL-4 in patients achieving disease stabilization compared with patients with progressive disease. Furthermore, patients were able to generate a positive ELISPOT response when stimulated with Lucanix specific antigen after vaccination.⁵⁷ These results indicated that favorable clinical response correlates with vaccine-induced immune responses justified further clinical investigation.

Table 2. Known TAAs of various cancers

Cancer origin	TAAs
Breast cancer	MUC 1, MUC 4,^99 Tn, T,^100 STn,^101 Her2/Neu,^102 BAGE, CAMEL, MAGE A1, Mammaglobin-A, TRAG-3, ML-IAP, Survivin, WT-1^103
Leukemia	hTERT, WT1, PRTN3, BCL2, LAMR1, PRAME, RHAMM/HMMR, G250/CA9^104
Lung cancer	MAGE-A3/4, NY-ESO-1, MAGE-A10, SOX2, CAGE, GBU4–5, Cyclin A, Cyclin B1, c-myc, STEAP, p53, Annexin 1, IMP1, p62/IMP2, CDK2, Survivin^103
Melanoma	MART, gp100, Tyrosinase, EBV BMLF1, CMV pp65,^105 BAGE, MAGE-A1/2/3/4/6/10/12, CAMEL, SSX2, TRAG-3, Melan-A, TRP1/2^103
Prostate cancer	PSA, PSMA, Eph2A, FGF-5, hTERT, AFP, GAGE-1/2/8, MAGE-A1, CAMEL, TRAG-3^103
Ovarian cancer	DAM 6–10, HER2/neu, hTERT, AIM-2^a, Cyp-B, ART-4, G250, Adipophilin^103
Pancreatic cancer	MUC 1, KRAS, EGFR,^84 Mesothelin,^106 p53^107

List of selected TAAs for breast cancer, leukemia, lung cancer, melanoma, prostate cancer, ovarian cancer, and pancreatic cancer.

The expression of costimulatory molecules is thought to increase the potency of allogeneic vaccines and direct the immune response. A phase I study enrolling 19 patients with advanced stage (stage IIIB, IV) metastatic NSCLC employed this strategy. Patients were vaccinated with an irradiated whole cell allogeneic vaccine (cell line AD100) transfected to express the costimulatory molecule B7-1 along with either HLA-A1 or HLA-A2.⁵⁵ The costimulatory molecule B7-1 was administered to boost the expansion of tumor specific immune responses that may help overcome the immunosuppressive environment present in NLSC. The patients were administered a vaccine depending on their HLA type; A1 patients received the A1 vaccine for direct antigen presentation, and non-A1 and non-A2 patients, received an unmatched vaccine (A1) for indirect antigen presentation. The median survival was found to be 18 mo with 32% of the patients achieving disease stabilization. After subgroup analysis, the authors found that disease stabilization was not HLA restricted, suggesting that indirect presentation not only sustained an anti-tumor response, but also enhanced it through allogeneic MHC molecules. Furthermore, an ELISPOT assay using CD8⁺ T-cells isolated from patients after treatment showed increased IFN-γ production by vaccine stimulated CD8+ T cells when compared with levels before treatment.⁵⁸ The ability of this allogeneic vaccine to stimulate a tumor specific CD8⁺ T-cell response and increase patient survival in patients with advanced disease bodes well for cancer immunotherapies.^55,58

Recently, results from a novel immunotherapeutic approach for the treatment of NSCLC indicate that allogeneic tumor cell lines secreting HSPgp96-Ig show that peptides from allogeneic tumor cells delivered to patients’ immune system via gp96 can induce an immune response.⁵⁹ While no objective tumor response was observed in clinical trials, stabilization of tumor growth and increased release of IFN-γ by CD8⁺ T Cells warrant further exploration of similar approaches.

Leukemia

Although chemotherapy, radiation therapy and transplantation have become the standard of care in leukemia, a study was conducted to test the efficacy of an allogeneic tumor lysate pulsed DC based vaccine in patients with B-cell chronic lymphoma leukemia (B-CLL).⁶⁰ The therapeutic regimen involved five intradermal administrations of DCs pulsed ex vivo with allogeneic tumor lysate or apoptotic bodies. Of the 9 patients enrolled in the study, 4 patients achieved disease stabilization for over 23 mo as determined by stabilized white blood cell and lymphocyte counts. However, only 1 of the 4 HLA matched patients was able to raise a detectable anti-TAA immune response. While this phase I trial shows that the therapy was tolerable and had positive immune correlates, immunotherapies in leukemia may provide maximum therapeutic benefit as an adjuvant therapy after reduction of disease burden through ablative therapies and transplantation. Thus, optimizing dosage of myeloablative drugs and immunotherapeutics may provide the greatest clinical benefit.

Chronic myeloid leukemia (CML) is another type of leukemia that responds well to chemotherapeutic drugs. However, the residual disease persists following treatment. A clinical trial was conducted to determine whether vaccination with GM-CSF-secreting K562 cells could induce an antitumor immune response in CML patients receiving the chemotherapeutic drug imatinib mesylate.⁶¹ The K562 cell line was chosen for this immunotherapy, as it is an established human CML cell line that expresses many CML antigens. Results from this clinical trial with 19 patients determined that vaccination with GM-CSF secreting K562 cells was capable of diminishing disease below detection levels in seven patients and lowered tumor burden in 12 patients. These results further emphasize the role of the immune system in eliminating and controlling the growth of residual cancer cells resistant to standard chemotherapies and the efficacy of immunotherapies when used in an adjuvant fashion.

Melanoma

The most notable melanoma immunotherapeutic to date, Canavaxin-polyvalent (PV), was tested at the John Wayne Cancer Institute in the largest single institution stage-III⁶² melanoma trial ever conducted. In the phase II trial, the allogeneic polyvalent whole cell vaccine formulation Canavaxin-PV, was demonstrated to be an effective adjuvant and immunotherapy. The vaccine regimen consisted of intradermal injections for 1 y following lymphadenectomy, and resulted in a significant increase in 5 y OS compared matched, untreated groups. However, it seems that the subgroup analysis, which identified Canavaxin-PV’s efficacy, did not correctly identify factors correlating to vaccine efficacy. The two subsequent randomized phase III trials testing Canvaxin-PV + the adjuvant BCG vs. placebo + BCG in stage-III and stage-IV melanoma were halted before completion, because no beneficial effect between the vaccine treated and the placebo treated group was observed.

While the reason for failure of this trial is uncertain, Canavaxin-PV had many characteristics needed for an allogeneic cell based vaccine. This formulation consisted of 3 melanoma cell lines (M10-V, M24-V, and M101-V) and contained over 20 distinct melanoma-associated antigens. Moreover, immunologic analyses of the phase II trial demonstrated a correlation between matched MHC haplotype and increased OS. This vaccine regimen, however, could be augmented by cytokine stimulation and the use of another adjuvant. This trial underscores the necessity for a multifactorial approach for effective immune activation in cancer patients.

Other treatment strategies for melanoma have focused on presentation of a number of TAAs through other methods. Established through a series of clinical trials, Melacine^® has exhibited the effectiveness of lysate based vaccines in the melanoma model.⁶³ The allogeneic melanoma cell lysate, utilized cell lines (Mel-D and Mel-S) isolated from biopsies of patients with nodular melanoma.⁶⁴ Administered as the lysate of multiple allogeneic cell lines with the adjuvant detoxified endotoxin and mycobacterium cytoskeleton (DETOX), this vaccine increased melanoma specific CTL precursors, increased cancer cell specific antibody titers, and induced immunity as evidenced by delayed-type hypersensitivity reactions in response to autologous cancer cells. In phase I clinical trials this vaccination regimen was able to induce remission in 5 out of 17 patients (29%).

Subsequent phase II and phase III clinical trials did not show such promising results, however, with only 5 complete remissions and 7 partial remissions out of 129 patients (overall objective response rate 6.1%). In an attempt to supplement the vaccination strategy, Melacine^® was administered along with a single dose of 300 mg/m² CY and IFN-α intravenously after 4 doses of Melacine^®. Of the 39 patients that were evaluated, there was only an overall objective response of 10%. Nevertheless, 64% of the patients achieved disease stabilization for 16 weeks.⁶⁵ Further investigation confirmed the induction of an anti-vaccine immune response against multiple melanoma antigens including Tyrosinase, MART-1, gp100, MAGE-1, MAGE-2, and MAGE-3.⁶⁶ The results of this trial augment a growing body of literature supporting the use of immunotherapies in a minimal residual disease setting.

While similar strategies such as immunization with recombinant MART-1 peptides have also been shown some efficacy, whole cell vaccines and cell lysates have the added advantage of stimulating T cells with unidentified intracellular and extracellular melanoma antigens.⁶⁷ Loading these antigens onto DCs confers the additional ability to directly prime CD8⁺ CTLs through cross presentation of exogenously acquired antigen. The efficacy of DCs to induce anti-tumor immunity was demonstrated in phase I clinical trials. Patients received autologous DCs loaded ex vivo with antigens from apoptotic/necrotic allogeneic melanoma cells. Anti-tumor immunity was confirmed by the induction of gp100 and MelanA/MART-1-specific CTLs.⁶⁸

This study supports data presented by Palucka et al. in 2006.⁶⁹ In this study, patients with stage-IV melanoma were vaccinated with autologous monocyte derived DCs loaded with killed allogeneic Colo829 melanoma cells and activated with GM-CSF, IL-4, TNF-α and CD40 ligand. The study participants had a median OS of 22.5 mo in conjunction with increased IFN-γ secretion by lymphocytes and proliferation in the presence of a Colo829 TAA peptide. CD8⁺ T-cell immunity to novel peptides was also observed, indicating that DC cross presentation had occurred.

The relative success of this phase I trial compared with allogeneic DC based vaccinations can be attributed to its multifactorial approach. The use of primed autologous DCs loaded with allogeneic TAAs from the Colo829 cell line, addressed problems concerning HLA matched presentation of antigen and DC anergy. This vaccination strategy demonstrates the importance of utilizing both the innate immune system and TAAs in eliciting an effective anti-tumor immune response.

Taken together, observations from these clinical trials in melanoma seem to indicate that polyvalent strategies that augment a sufficient immune response need to be considered.

Prostate Cancer

The development of an effective immunotherapy to treat prostate cancer has seen modest success in recent years. The immunotherapy Onyvax-P consists of three irradiated prostate cancer cell lines (OnyCap23, LnCaP, and P4E6) representative of three different stages of prostate cancer. The vaccine formulation was administered with the adjuvant BCG to individuals with asymptomatic hormone-resistant prostate cancer (HRPC). Among the 26 patients with increased levels of PSA enrolled in the study, administration of the vaccine resulted in decreased PSA velocity (PSAV) in 42% of the patients and increased median time to disease progression from 28 weeks to 58 weeks.⁷⁰ Immunological studies indicated that patients with decreased PSAV showed a Th1 cytokine release pattern when T cells were stimulated by lysate ex-vivo compared with patients who did not respond to the vaccine and exhibited both a Th1 and Th2 cytokine release patterns.

The subsequent phase I/II clinical trial explored the administration of a whole cell vaccine consisting of a mixture of three proprietary (Onyvax Ltd.) prostate cancer cell lines⁷¹ along with the bacterial adjuvant BCG.⁷² Throughout the course of the study, levels of PSA, cytokines, prostate-specific T cells, and serum antibody titer were measured. Although small decreases in PSA could not be attributed solely to the vaccine, increased levels of cytokine production and prostate-specific antibody titer indicated that the vaccination led to a specific immune response. This series of clinical trials highlights the importance of disease burden on the efficacy of an allogeneic tumor immunotherapy approach. The initial vaccine trial for Onyvax-P enrolled patients with advanced HRPC and showed limited efficacy. However, the successful treatment of patients with asymptomatic HRPC in this study indicates the limitation of solely administering immunotherapies to patients with high disease burdens.

A cytokine secreting allogeneic modality was tested in prostate cancer by the GVAX platform. The regimen consisted of two allogeneic cell lines (PC3, LNCaP) modified to secrete GM-CSF.⁷³ PC3 is derived from prostate cancer bone marrow metastases and is castration resistant.⁷⁴ The LNCaP cell line derived from prostate cancer lymph node metastases is hormone sensitive, and expresses numerous prostate-associated antigens.⁷⁵ Together, these two cells lines provide a broad range of antigens that are prevalent in various prostate cancers. Preclinical data showed that GVAX was successful in inducing anti-tumor immunity in melanoma, lymphoma, colon, fibrosarcoma, lung, renal, and prostate cancers. A potential mechanism for this anti-tumor response is the concurrent activation of APCs and arms of the adaptive immune system.⁷⁶ Phase I/II clinical trials testing the efficacy and safety of GVAX as a monotherapy, enrolled 34 patients with metastatic castration resistant prostate cancer (CRPC). The trial demonstrated that patients receiving the vaccine had a median OS of 26.2 mo, comparable to patients that receive chemotherapy.⁷⁷ Based on the findings of these small phase I/II trials, two large Phase III randomized studies (VITAL-1 and VITAL-2) were developed to test the clinical efficacy of GVAX in treating metastatic prostate cancer.⁷⁸

Production of GVAX was halted after both VITAL-1 and VITAL-2 were terminated due to the inefficacy of GVAX and excessive deaths, respectively.⁷⁹ Understanding the failure of GVAX in metastatic prostate cancer would provide useful information toward the improvement of this vaccination and similar modalities.

To test GVAX in a combinatorial setting, a clinical trial was performed in which both GVAX and Ipilimumab were administered in escalating doses.⁸⁰ Ipilimumab is a fully human IgG CTLA-4 blocking antibody. CTLA-4 is a homolog of CD28 found on T cells generally 2–3 d post activation. Its ability to bind B7-1 and B7-2 with a 100-fold greater affinity than CD28 allows it to compete with CD28 for B7-1 and B7-2 binding, thus terminating the T cell response and limiting the T cell population pool size.⁸⁰ The study showed disease regression as determined by reduction of PSA levels in 5 of 12 patients diagnosed with metastatic HRPC. This correlated directly with Ipilimumab dosage as 5 of the 6 patients received a higher dosage of Ipilimumab.⁸¹ Although GVAX was more efficacious as a combinatorial treatment, patients faced several endocrine complications at higher dosages indicating the need for more specific targeting mechanisms if such combinatorial therapies are to be employed.

Similar to GVAX, the LNCaP cell line was modified using a retroviral vector to secrete IL-2 and IFN-γ.⁸² Six patients with HRPC were enrolled in the study. During the course of the study serum PSA levels and CTL generation were monitored. A 2-fold higher response to various antigens was seen in all patients. Interestingly, an inverse correlation was observed between levels of serum PSA and antigen specific immune cells in circulation. These results support the hypothesis that T cells migrate from the periphery into vascular circulation to mediate an immune response.

As these clinical trials in prostate cancer have demonstrated, a vaccine response can vary widely based on cytokine costimulation, allogeneic cell lines administered and vaccine modality. Moreover, these trials also identify disease burden as a significant factor in the efficacy of an immunotherapy during clinical evaluation. It is thus critical to determine the appropriate clinical settings in which these immunotherapies are most effective as a monotherapy, combinatorial therapy, or as an adjuvant therapy.

Pancreatic Cancer

Since 1975, the rate of incidence of pancreatic cancer has been increasing by 1.5% per year, a figure attributed partly to the lack of a widespread detection method.⁸³ Although TAAs (Table 2) have been identified for pancreatic cancer, correlation between detection of these antigens and disease prognosis remain highly unreliable. This lack of knowledge justifies the exploration of allogeneic whole cell vaccines as immunotherapies for patients with pancreatic cancer.⁸⁴

Currently, the standard of care for advanced pancreatic cancer is the chemotherapeutic drug gemcitabine.⁸⁵ Clinical trials of allogeneic vaccines for pancreatic cancer are currently underway and the early phase of trials have reported promising results. Two allogeneic vaccines in currently tested in clinical trials for the treatment of pancreatic cancer include NewLink Genetic’s HyperAcute® pancreatic cancer vaccine (Algenpantucel-L) and BioSante Pharmaceutical’s GVAX Pancreas Vaccine. Algenpantucel-L (HyperAcute-Pancreas) combines two human allogeneic pancreatic cancer cell lines expressing α-1,3-galactosyltransferase, a murine enzyme that mediates α-galactosyl (αGal) epitope expression on the surface of the cells. The rationale behind this vaccine strategy is to take advantage of naturally produced, anti-αGal antibodies by the human body as an adjuvant.⁸⁶ The anti-tumor immune response resulting through opsonization, complement activation, and ADCC of anti-αGal antibody:αGal immune complexes.^87,88 The phase II clinical trial for HyperAcute-Pancreas generated favorable results with 62% of patients remaining disease-free at the one-year endpoint, and an overall survival of 24.1 mo.⁸⁹ A phase III clinical trial, which began in 2010, enrolled 722 patients with stage-I and stage-II pancreatic cancer to study the effect HyperAcute-Pancreas in combination with chemotherapy or chemoradiotherapy.

The previously discussed GVAX vaccine platform has also been tested in pancreatic cancer. In an initial 14-patient phase I clinical trial testing for the safety of the GVAX regimen, Jaffe et al. reported that 3 of 14 patients were disease-free for at least 25 mo after treatment.⁹⁰ Combinatorial approaches employing GM-CSF secreting cell lines and chemotherapy are also being explored in metastatic pancreatic cancer. In 2008, a phase I clinical trial purported minimal toxicity and enhanced T cell function in patients receiving an immunotherapy and cyclophosphamide.⁹¹ These findings suggest chemotherapeutic treatments can enhance the activity of an immunotherapy as exhibited by higher rates of antigen-specific CD8⁺ T-cell activation.

Another clinical trial for a GM-CSF secreting immunotherapy also recently explored the combinatorial use of immunotherapy, radiotherapy and chemotherapy as an adjuvant, post resection of pancreatic adenocarcinoma.⁹² This single arm phase II trial demonstrated that administration of up to 5 doses of the vaccine induced antigen-specific T cells correlating with increased overall survival. Patients that remained disease free generated lymphocytes that could respond to a greater variety of antigen after receiving the combinatorial therapy. The results of these trials provide a strong foundation for growing evidence that immunotherapies should be considered for use as an adjuvant therapy to resection or in a combination with established chemotherapeutic regimen in the treatment of pancreatic cancer.

Other Cancers

The CancerVax regimen designed initially for the treatment of melanoma, consists of three live-irradiated melanoma cell lines that include many immunogenic colon carcinoma TAAs including the glycoprotein TA90. A phase I clinical trial was conducted to determine whether administration of CancerVax and the adjuvant BCG could induce a vaccine-specific immune response which correlated with overall survival in patients with advanced colon cancer.⁹³ The trials results indicated that vaccination increased median OS compared with historical controls and that patients with high titers of TA90-IgM immune complexes had a significantly higher OS. This trial demonstrates the potential for allogeneic therapies to treat a broad array of cancers given a set of common TAAs.

In a clinical trial for metastatic renal cancer, 10 HLA-A-0201 matched patients were enrolled to test the efficacy of an HLA matched allogeneic renal cancer cell line (RCC26) transfected with IL-7 and B7-1.⁹⁴ Despite an increase in median OS, immune correlates suggested that vaccination induced a Th2-polarized response marked by release of IL-10 upon stimulation with vaccine-specific peptide. Furthermore, the vaccine failed to rescue tolerized lymphocytes, which were incapable of interferon production both in the presence of antigen-specific and non-specific stimulation. This implies that when employed as a monotherapy, the vaccine was not able to break tolerance. Therefore, this therapy must be employed either, concomitantly to deplete Tregs and break tumor immune suppression, or in a minimal residual disease setting.

The Future of Therapeutic Allogeneic Cancer Vaccines

These clinical trials demonstrate the ability of cell based allogeneic immunotherapies to stimulate a clinically relevant response. The presence of a tumor mass within an established immunosuppressive microenvironment is often an insurmountable barrier, for the immune system to overcome. As demonstrated by the vaccination of individuals with resected lung cancer and asymptomatic HRPC, the use of immunotherapies in a clinical setting either post surgery or post chemotherapy allows the immune system to respond more efficiently by killing residual tumors that escape initial treatment.¹⁴ Furthermore phase I studies, which are performed on patients with advanced stage cancers with a poor prognosis provide a suboptimal environment for the assessment of these immunotherapies. Complicating these results further, patients participating in many clinical trials are often severely immunocompromised due to prior participation in various chemo and radiation therapies. With these limitations of the clinical trials process, allogeneic cancer vaccines could show the most clinical benefit when used as an adjuvant therapy.⁹⁵

It is also imperative to note the correlation between the effectiveness of an allogeneic vaccine and the number of common TAAs (Table 2) expressed by both the cancer and the allogeneic cell line. The discovery of multiple TAAs that serve as excellent biomarkers for the disease allows not only for easy monitoring, but also for better engineering of cancer cell lines and the induction of anti-tumor specific immunity. This has been well documented in the literature, as vaccination with a greater number of allogeneic cell lines has proven more effective.

A major impediment to the use of allogeneic therapies, however, is the extensive intratumoral heterogeneity in cancers.⁹⁶ Through multi region tumor sequencing, one study has determined that between 63% and 69% of all somatic mutations were heterogeneous.²⁰ Additionally, the study enabled the construction of gene phylogenies, which unveiled the progressive loss of tumor-suppressor genes within a single tumor, indicating the convergence of tumor phenotype through multiple spatially separated mutations.²⁰ This line of evidence supports a model in which following the initial mutations needed for tumor escape, subclones are free to mutate divergently in a deregulated manner.⁹⁷ Moreover, the addition of selective pressures through the use of allogeneic immunotherapies may promote escape and drive further mutation and divergence. Thus it can be hypothesized that allogeneic tumor cell vaccines developed from established cell lines may not represent the antigenic characteristics of the entire tumor and therefore will face increasing levels of intratumoral complexity depending on the tumor progression within a given patient. Despite the induction of measurable antitumor immunity in many patients, the lack of successful remission or significant clinical benefit of allogeneic tumor immunotherapies when employed as a monotherapy may be explained by the pre-existing inter- and intra-tumoral heterogeneity and/or tumor induced immunosuppression at the microenvironment. In addition to the challenge posed by the inter- and intra-tumoral heterogeneity, many of the allogeneic tumor cell vaccine clinical trials discussed above did not incorporate strategies that will mitigate the tumor-induced immunosuppression. Therefore, in the future, a better therapeutic efficacy is expected when allogeneic tumor vaccines are developed based on the genomic and transcriptomic analysis of patient’s cancers⁹⁸ and administered in conjunction with anti-immunosuppressive agents.

Abbreviations:
APC	=	antigen presenting cell
B-CLL	=	B-cell chronic lymphoma leukemia
CIK	=	cytokine induced killer
CRPC	=	castration resistant prostate cancer
HSP	=	heat shock protein
CTLs	=	cytotoxic T lymphocytes
CY	=	cyclophosphamide
DC	=	dendritic cell
DOX	=	doxorubicin
GM-CSF	=	granulocyte macrophage-colony stimulating factor
GVHD	=	graft versus host disease
GVL	=	graft versus leukemia
HRPC	=	hormone-refractory prostate cancer
MHC	=	major histocompatibility complex
NK	=	natural killer
NSCLC	=	non-small cell lung cancers
OS	=	overall survival
PFS	=	progression free survival
PS	=	phosphatidyl serine
PSA	=	prostate-specific antigen
TAAs	=	tumor-associated antigens

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Acknowledgments

This work was supported by NIH grants R01 CA138993 (to Selvaraj P), F31 CA165632 (to Patel JM), and F31 CA165897 (to Bozeman EN). The authors thank Archana V Boopathy for critical reading of the manuscript.

10.4161/hv.26568