ABSTRACT: The availability of several novel antibodies, coupled with viral, DNA, and dendritic-cell vaccines, has renewed interest in immunotherapeutic approaches to the treatment of advanced prostate cancer. Although promising, none of these approaches have led to major clinical activity, and in the case of cell-based immunotherapy with GVAX, new concerns about safety arose when this therapy was used in the castration-resistant setting. A more attractive yet toxic approach has also utilized a check-point blockade with CTLA-4 antibodies. Although initial clinical efficacy has been observed, toxicity appears to be the major limitation of its use in prostate cancer. Sipuleucel-T (Provenge) is an autologous active cellular immunotherapy product that includes autologous dendritic cells pulsed ex vivo with PAP2024, a recombinant fusion protein made of prostatic acid phosphataase (PAP) and granulocyte–macrophage colony-stimulating factor (GM-CSF). Despite the lack of objective anti-tumor activity seen with sipuleucel-T, a recently reported phase III trial demonstrated a significant improvement in the overall survival of men with asymptomatic, minimally symptomatic metastatic castration-resistant prostate cancer (CRPC). This agent is the first FDA-approved novel immunotherapeutic compound for the treatment of a solid malignancy. A better understanding of how clinicians should incorporate this novel agent into the current management of CRPC is needed.
Prostate cancer is the second-leading cause of death in men in the United States; more than 217,730 new cases were expected to be diagnosed in 2010. Although the majority of patients with advanced prostate cancer have an initial response to androgen deprivation, essentially all patients eventually progress to a castration-resistant state, manifested by rising levels of prostate-specific antigen (PSA), progressive disease on imaging studies, and/or worsening of symptoms—all in the setting of an anorchid testosterone level. Historically, patients were offered a variety of secondary hormonal manipulations followed by palliative chemotherapy with docetaxel(Drug information on docetaxel) (Taxotere)-based regimens. Following a long period in which there were limited therapeutic developments, the FDA recently approved cabazitaxel (Jevtana), a second generation taxane for patients with disease progression following docetaxel. In addition, recently presented evidence of improved survival for patients receiving the selective CYP-17 inhibitor abiraterone following docetaxel-based chemotherapy will likely result in the approval of another agent in this setting.
A better understanding of the role of inflammation and immune activation in prostate cancer, coupled with the long natural history of the disease, its ability to induce auto-antibodies, and the availability of several tissue-specific proteins that may serve as prostate tumor antigens,[6,7] has facilitated the development of several immune-based therapeutic strategies. In this review, we summarize recent clinical developments in prostate cancer immunotherapy, and we specifically review the clinical development of sipuleucel-T (Provenge), a novel immunotherapeutic agent recently approved by the FDA for the management of asymptomatic or minimally symptomatic metastatic castration-resistant prostate cancer (CRPC).
Granulocyte–macrophage colony-stimulating factor (GM-CSF) is a small secreted polypeptide that binds to specific cell surface receptors; it is a potent pleiotropic cytokine capable of enhancing hematopoietic differentiation and activation from several lineage precursors, including phagocytic macrophages and dendritic cells. This non-specific activation of antigen-presenting cells (APCs) results in increased co-stimulatory molecule expression on host bone marrow–derived APCs. When administered in vivo, GM-CSF promotes the growth and antigen-presenting capabilities of dendritic cells, leading to T-cell cross-priming. The therapeutic effect of recombinant GM-CSF as an anticancer agent has been studied in several models. Injection of a murine tumorigenic T-leukemia cell line expressing mGM-CSF into pre-established tumors of syngeneic mice led to a significant regression of the tumors. Additionally, mice injected with melanoma cells transfected with a recombinant GM-CSF gene either completely rejected the tumor cells or developed tumors with a mean volume 50 times smaller than in controls. Furthermore, vaccination with GM-CSF-expressing MPC11 cells has been shown to induce a potent antitumor cytotoxic T-lymphocyte response associated with tumor rejection in the majority of the animals tested. Additional studies of autologous or allogeneic prostate tumor cells transfected to produce GM-CSF have demonstrated the generation of potentially relevant immune responses.[11,12]
Exogenous administration of GM-CSF has been studied in men with advanced prostate cancer.[14-16] Although the dose and schedule of administration has varied among clinical trials, it is clear that GM-CSF administration can lead to declines in PSA levels in patients with hormone-naïve and castration-resistant disease. However, these declines are often transient, since disease eventually progresses either serologically or radiographically in all patients. The addition of maintenance therapy to GM-CSF also appears to increase the rate of PSA reduction, with no evidence that GM-CSF interferes with PSA measurement or secretion. Although the clinical significance of such PSA changes remains uncertain, these data suggest a potential biologic effect of GM-CSF on PSA. This observation alone is important, since many vaccination strategies have used GM-CSF as an adjuvant. However, the confounding effect that GM-CSF may have on PSA levels makes interpretation of studies that utilize GM-CSF somewhat problematic.
In recent years, cytotoxic T-lymphocyte (CTL)-associated antigen-4 (CTLA-4)-targeted therapies have undergone the most extensive evaluation in multiple solid tumors, including prostate cancer. T-cell activation is dependent on the ability of the T-cell receptor to recognize specific antigen-bound major histocompatibility complex (MHC) peptides. Additional antigen-independent costimulatory signals are required for the generation of a T-cell response. These include the normal interaction between CD28 and CTLA-4, both costimulatory receptors located on the surface of T cells, and costimulatory molecules B7-1 (CD80) and B7-2 (CD86), present on APCs. While ligand engagement by CD28 activates T cells, the binding of CTLA-4 and B7 sends an inhibitory signal that down-regulates T-cell activation and serves as a natural brake, altering downstream cytokine production and cell-cycle proliferation, both required for growth.[17-19] In vivo inhibition of CTLA-4 and its ligands through the use of a neutralizing antibody has been shown to induce antitumor activity in several murine tumor models.[20-22] Therefore, blockade of CTLA-4 may represent an important mechanism for potentiating T-cell immunity and anti-tumor response.
Ipilimumab and tremelimumab are two humanized anti-CTLA-4 antibodies that are the lead compounds in the investigation of this immune approach in solid tumors. Small et al evaluated ipilimumab both as monotherapy and in combination with GM-CSF in men with CRPC.[23,24] In the monotherapy trial, 14 patients received ipilimumab as a single dose of 3 mg/kg IV. Treatment was well tolerated, with clinical autoimmunity limited to one patient who developed grade 3 rash/pruritus that required systemic corticosteroids. Although a trend for an increasing percentage of CD4 and CD8 T cells co-expressing MHC class II was observed, flow cytometry analysis of lymphocyte sub-populations failed to demonstrate polyclonal T-cell activation. A subsequent study evaluated the combination of CTLA-4 blockade and GM-CSF in 24 patients with CRPC. While GM-CSF was given as a fixed dose of 250 μg/m²/day for 14 days, ipilimumab was dose-escalated from 0.5 mg/kg to 3 mg/kg, given once every 4 weeks in a sequential phase I design. Observed toxicity was similar to that seen in previous CTLA-4 studies, with 4 patients developing grade 3 adverse events, including a cerebrovascular accident, skin rash, pan-hypopituitarism, temporal arteritis, and diarrhea. Of interest, two of these patients were among those who obtained the greatest clinical benefit, as assessed either radiographically or by PSA level. Immune-based studies have also suggested the enhancing effects of this combination on the activation of circulating CD8+ T cells. Although disappointing clinical results have been observed in other solid tumors in studies using similar antibodies, the recent overall survival (OS) results from a phase III study of ipilimumab with or without a gp100 peptide vaccine in patients with metastatic melanoma clearly suggest that when the appropriate antigen is selected, antibody-mediated therapy has significant potential to become an an important therapeutic strategy in solid malignancies. Although the ideal antigen for prostate cancer immunotherapy has not been well defined, the clinical experience with ipilimumab in melanoma provides impetus for continued immune-based research in prostate cancer.
DNA, Viral, and Cell-Based Vaccines
Several vaccine approaches have undergone clinical evaluation in men with prostate cancer.
One approach involves DNA-based constructs that ultimately activate a specific antitumor immune response. An advantage of this approach is that PAP is a unique antigen in prostate cancer cells, and unlike other vectors (eg, virus), DNA can be rapidly and precisely synthesized. When contrasted with viral vectors, one of the drawbacks of using DNA as a vector is its low level of immunogenicity. Initial studies in rat models using large and repetitive doses of a GMP-grade plasmid DNA vaccine encoding human PAP (pTVG-HP) demonstrated the ability of this vaccine to elicit effective PAP-specific CD4 and CD8 T cells. In a subsequent phase I/II study, 22 prostate cancer patients with biochemical recurrence-only disease were treated in a dose-escalation trial with 100 μg, 500 μg, or 1,500 μg of TVG-HP plasmid DNA, co-administered intradermally with 200 μg of GM-CSF as a vaccine adjuvant, six times at 14-day intervals. Although no major changes in PSA levels were reported, treatment was well tolerated, with minimal if any adverse events, and a significant PAP-specific CD4+/CD8+ T-cell proliferation that included interferon gamma(Drug information on interferon gamma) (IFN γ)-secreting CTL activity was observed.
Using viral or bacterial vectors as a vehicle for vaccine delivery is a cost-effective way to allow the insertion of multiple genes for tumor-associated antigens. Of note, some of these vectors can cause direct local inflammation, which in turn can increase APC recruitment and antigen presentation. Poxviruses have been employed as a vehicle for prostate cancer vaccines in various combinations. Vaccination with recombinant vaccinia virus expressing human PSA (rV-PSA) has been shown to be safe and to induce PSA-specific T-cell responses in animal models and in humans.[30-32]
A phase II trial of the Eastern Cooperative Oncology Group (ECOG) evaluated a heterologous prime/boost vaccination schedule with vaccinia and fowlpox viruses expressing human PSA, in men with BCR-only disease. Patients were randomly assigned to receive four vaccinations with fowlpox-PSA (rF-PSA), three rF-PSA vaccines followed by one rV-PSA vaccine, or one rV-PSA vaccine followed by three rF-PSA vaccines. The primary endpoint was PSA response at 6 months. There were minimal adverse events and no autoimmune events, and almost half of the patients (45.3%) remained free from PSA progression at 19.1 months of clinical follow-up. An increase in PSA-specific T-cell precursors was observed in 46% of patients; however, none of the men in the study had evidence of an anti-PSA antibody response.
PROSTVAC, a construct of fowlpox and vaccinia vectors that contains a triad of costimulatory molecule transgenes (intercellular adhesion molecule 1, B7-1, and leukocyte function associated antigen 3) has also been studied in prostate cancer. Two phase I studies demonstrated the safety and immune effects of this compound in patients with CRPC.[34,35] In a subsequent randomized double-blind phase II study, 125 patients with minimally symptomatical metastatic CRPC were randomly assigned in a 2:1 fashion to receive either PROSTVAC-VF plus GM-CSF or empty vectors plus saline injections. Although no difference in the primary endpoint of progression-free survival (PFS) was found (3.8 vs 3.7 months; hazard rate [HR], 0.88; P = .60), patients treated with PROSTVAC-VF had a superior OS compared with the OS of those in the control arm (25.1 vs 16.6 months; HR, 0.56; P = .0061). The toxicity was similar to that seen with other viral vectors tested in prostate cancer. Grade 1 and 2 injection site reactions, fatigue, fever, and nausea were the most common adverse events reported. No detectable antibody response to PSA was observed, and all patients had an augmented antibody response to the vaccinia and fowlpox vectors. There was no association between the antivector antibody production and OS. A pivotal phase III trial of PROSTVAC-VF is currently planned.
The use of allogeneic prostate cancer cells as immunotherapy vectors has been another attractive immunotherapeutic approach in advanced prostate cancer. GVAX is a combination of two prostate carcinoma cell lines (LNCap and PC-3) modified with the GM-CSF gene and thereby engineered to increase the antigen-presenting capabilities of dendritic cells to T cells.[37-40] Two subsequent phase II studies evaluating the clinical activity and safety of GVAX in men with CRPC[41,42] demonstrated an improvement in the median OS compared with historical data from the Halabi nomogram, a validated pre-treatment prognostic model for CRPC. In the first trial, Small et al reported that the median OS for GVAX-treated patients was 26.2 months, compared with the 19.5 months (P = .01) predicted by the nomogram. The trial by Higano et al, which employed a vaccine construct that was re-engineered to secrete a higher dose of GM-CSF, also demonstrated a superior OS benefit (35 months) compared with the same historical control. When the immune effects of treatment were evaluated, immunoreactivity was found not to be a significant predictor of survival or clinical response. The most common adverse events reported in both studies were injection site reactions and flu-like symptoms.
On the basis of these results, two large randomized phase III trials were launched. The first (VITAL-1) was a head-to-head comparison of GVAX vs docetaxel/prednisone, and the second (VITAL-2) was a combination of docetaxel and GVAX vs docetaxel/prednisone. Although VITAL-1 completed its target accrual of 600 patients with CRPC, VITAL-2 was halted after an early interim analysis demonstrated an imbalance in the number of deaths in the immunotherapy arm (67 vs 47 in the docetaxel/prednisone arm). On the basis of this finding, an interim analysis for VITAL-1 was conducted. There was no OS difference between GVAX and docetaxel/prednisone (20.7 and 21.7 months, respectively; HR, 1.03; P = 0.78). To date, the relationship between GVAX and docetaxel treatment and deaths remains unknown. Based on these results, further clinical development of GVAX for prostate cancer has been terminated.