Immunological therapy in urological malignancy: Novel combination strategies
David S Finley md, Institute of Urologic Oncology, David Geffen School of Medicine at UCLA, 924 Westwood Blvd, Suite 1050, Los Angeles, CA 90024, USA. Email: firstname.lastname@example.org
At present, immunotherapy in urological malignancy is experiencing a renaissance, particularly with the emergence of a host of innovative cancer vaccines. Herein, we will review promising immunotherapeutic approaches and evaluate the data supporting their inclusion in novel combination strategies.
Spontaneous regression of malignancy has been observed on rare occasions in patients with renal cell carcinoma (RCC), bladder cancer, prostate cancer, melanoma and breast cancer, for example. This phenomenon is thought to occur as a result of immune mediated mechanisms.1–8 In RCC, spontaneous regression has been reported most frequently in the setting of metastatic disease, usually associated with treatment of the primary lesion.9–13 Various immunological mechanisms have been proposed to account for this association.14 The first link between the immune system and cancer was established in the late 1800s when a German surgeon, Dr Friedrich Fehleisen, identified the link between erysipelas and Streptococcus pyogenes and began treating cancer patients with bacterial cultures to induce tumor regression.15 This phenomenon was later recapitulated by William B Coley using a vaccine derived from Streptococcus pyogenes and Serratia marcesens (Coley's toxins) to innoculate cancer patients.16,17 Coley would later become known as the “father of immunotherapy” based on his seminal work with bacterial toxins. One hundred and thirty years later, an increased understanding of the immune system, the processes of tumor immune surveillance and immunoediting, and mechanisms of tumor escape have fostered a transition from non-specific “shot-gun immunotherapy” to rationally designed approaches with novel delivery mechanisms resulting in increased efficacy and specificity with decreased side-effects.
Despite an increased understanding of the molecular, genetic and epigenetic mechanisms underlying these cancers, the clinical efficacy of most immunotherapeutic agents as stand-alone treatments have yet to meet their promise – durable partial responses are infrequently enjoyed and complete responses have rarely been achieved. Numerous biological barriers exist to explain the limited antitumoral efficacy of existing strategies, including insufficient target specificity and individual patient personalization, paucity of target antigens, inability to induce robust immune activation, tumor induced immunosuppression of antigen presenting cells, ineffective delivery of treatment agents to tumor sites, lack of tumor penetration and uptake as a result of vascular, stromal, interstitial and distribution heterogeneity within tumors, and various tumor cell escape mechanisms from immune surveillance.18 Perhaps most importantly, immunotherapy trials have traditionally enrolled patients with late stage, refractory disease. In the setting of a bulky disease, it seems unlikely that a monotherapy could overcome such a tumor burden by itself.
In this article, we review promising immunotherapeutic approaches in urological malignancy and some of the barriers to their success. In addition, we suggest several strategies for immune boosting using novel combinations of existing therapies based on emerging data.
Conventional immunotherapy and targeted therapy in RCC
Immunotherapeutic techniques for RCC were developed in the 1960s based on adoptive transfer principles. In early clinical trials, a survival advantage was shown for patients with metastatic disease (mRCC) using a purified RNA extracted from sheep immunized with human RCC cells.19 Interferons were the first cytokine studied in patients with mRCC. Initial data published in 1983 showed responses in up to 26% of patients treated with Interferon-alpha (IFN-α).20 Subsequent trials using both interferons as well as interleukin-2 (IL-2) showed variable results. A meta-analysis of 6117 patients extracted from more than 50 randomized trials found an overall chance of partial or complete response to immunotherapy of 12.4%.21 A survival advantage was also found when immunotherapy was utilized in conjunction with cytoreductive nephrectomy.22 The combination of IFN-α and nephrectomy in patients with mRCC resulted in a 10-month survival advantage compared with interferon alone in one study.23 The combination of interferon 2b and cytoreductive nephrectomy also conferred a survival benefit, although to a lesser extent, reporting a difference of approximately 3 months.24 A combined analysis of two of the largest prospective randomized trials of cytoreductive nephrectomy and immunotherapy, SWOG 8949 and EORTC 30957, showed a median survival of 7.8 months with IFN-α alone compared with 13.6 months for IFN-α in combination with cytoreductive nephrectomy.25
Although complete responses are rare with cytokine therapy, several studies have observed this phenomenon with both interferon and IL-2.26 For example, among patients with mRCC, most of whom underwent prior nephrectomy, weekly subcutaneous administration of pegylated interferon-2b was associated with a 3% complete response rate, 28% partial response and 47% stable response; progression-free survival was 5 months and median overall survival was 31 months.27 Another study showed a complete response rate of 7% in patients treated with high dose IL-2.26 Combination therapies with IFN-α and low dose IL-2 have shown activity as well, with responses correlating with expression of molecular markers, such as Ki-67 and Bcl-2, in the primary tumor.28
Although cytokine therapy with IL-2 has been the mainstay therapy capable of inducing durable remission, nephrectomy in combination with targeted therapies, such as sunitinib, sorafenib and temsirolimus, have shown significant activity.29–31 Motzer et al. reported improved progression free survival and response rates with sunitinib compared with IFN-α in patients with mRCC.30 A pooled analysis of two multicenter phase II trials of sunitinib in cytokine refractory mRCC showed stable disease in 24% of patients and a partial response rate of 42%.32 A randomized trial of temsirolimus, interferon or the combination of both agents showed an overall survival advantage for temsirolimus in patients with advanced kidney cancer.29 Other combination strategies with activity have included the use of IFN-α and active vitamin D3, IL-2 in combination with autologous lymphokine-activated killer cells (LAK) cells or tumor-infiltrating lymphocytes in patients with mRCC.33–35 Although the various combinations of drugs and surgery have shown a modest degree of activity, a common drawback to these approaches has been a lack of specificity.
Cancer vaccines are designed to elicit a specific anti-tumor response to tumor antigens to stimulate cell mediated and humoral immune responses. Two general categories of cancer vaccines exist: tumor-antigen specific vaccines, which are designed to deliver peptides or proteins; and tumor-cell vaccines, in which autologous or allogeneic tumor cells are modified to secrete cytokines or other immunostimulatory molecules capable of recruiting antitumor effector cells. This latter category includes tumor cell lysates, peptides and whole-cell based vaccines (e.g. inactivated autologous tumor cells [i.e. irradiated], gene modified tumor cell-based and dendritic cell [DC]-based vaccines). DC-based vaccines are thought to hold the most potential, given the central role of DC as powerful antigen presenting cells (APC) and “gatekeepers” for initiating and maintaining immunity.36
Dendritic cell-based vaccines
Various strategies have been used to induce a specific antitumor response by isolating DC, loading them with various tumor antigens or peptides, and then giving them back to the patient as a “Trojan-horse” vaccine. The recently FDA approved DC-based vaccine, Provenge (sipuleucel-T), is an autologous vaccine consisting of ex vivo loaded DC with the fusion protein prostatic acid phosphatase-granulocyte-macrophage colony-stimulating factor (PAP-GMCSF). In hormone refractory metastatic prostate cancer (HRPC), Provenge was used in the first randomized, double-blind, placebo-controlled trial of an autologous vaccine, reporting a median survival benefit of 4.3 months (hazard ratio HR 1.5, P = 0.01).37 A more recent randomized double-blind study administered a pox-based vaccine called PROSTVAC-VF.38 PROSTVAC-VF, by contrast, is a prostate-specific antigen (PSA)-recombinant vaccinia vector combined with a triad of immunostimulatory molecules (B7.1, ICAM-1, LFA-3) and subsequent boosts of rF-PSA-TRICOM (fowlpox boosting agent) and GM-CSF adjuvant. Among patients with HRPC, PROSTVAC-VF established a median overall survival benefit of 8.5 months.38 One might speculate that the improved efficacy of this vaccine was related to more robust adjuvant stimulation compared with Provenge.
For metastatic RCC, a DC-based vaccine similar to Provenge, linking GMCSF-CAIX, is currently under investigation. Carbonic anhydrase IX (CA-IX), is a renal tumor antigen involved in the hypoxic response that is highly expressed in clear cell RCC.39 Preclinical work showed that adenoviral mediated transduction of monocytes with a recombinant GMCSF-CAIX construct could stimulate CA-IX specific T cell mediated immune responses against RCC cells expressing CAIX and inhibit tumor growth.40 In a subsequent study, human DC were successfully transduced with the GMCSF-CAIX gene, resulting in production of a GMCSF-CAIX fusion protein that was capable of inducing DC maturation and CAIX specific cytotoxic T lymphocyte (CTL) induction.41 Recently, several interesting DC-based strategies have emerged. Yang et al. at UCLA developed a lentivirus vector system containing a transgene for a glycoprotein called SVGmu, which targets the surface protein, Dendritic Cell-Specific Intercellular adhesion molecule-3-Grabbing Non-integrin (DC-SIGN).42 This type of system with DC-targeting specificity represents a promising avenue for the selective delivery of engineered immunogens to DC to stimulate an antigen-specific immune response. Several other DC based approaches are under investigation.43,44
Others have attempted to exploit the specificity of CAIX using peptide vaccines. Shimizu et al. identified a series of CAIX derived peptides that have the ability to induce HLA-A24-restricted (the predominant HLA type in Japanese patients), tumor-specific CTL activity against RCC.45 Uemura et al. then injected 23 HLA-A24 positive patients with metastatic cytokine refractory CAIX positive RCC and evaluated tumor response by Response Evaluation Criteria In Solid Tumors and surrogate immunological parameters consisting of peptide-specific CTL response (IFN-γ release assay) and peptide specific IgG production.46 Successful induction of peptide specific CTL was shown, as well a production of anti-CA9p323 IgG; despite favorable immune response parameters, a clinically apparent partial response was only seen in three patients.
In bladder cancer, a phase I trial in post-cystectomy patients is underway using a fusion protein consisting of a monoclonal antibody with specificity for the DC receptor, DEC-205, linked to the highly immunogenic cancer-testis antigen, NY-ESO-1 (A study of CDX-1401 in patients with malignancies known to express NY-ESO-1. URL http://clinicaltrials.gov/ct2/show/NCT00948961?term=1401&rank=2[accessed on September 1, 2010]). In addition, a phase II trial is underway using oncogenic MPHOSPH1 and DEPDC1 derived peptides in conjunction with bacillus Calmette–Guérin (BCG) in patients after transurethral resection of bladder cancer with no sign of residual disease (W Obara et al., pers. comm., 2010). Considering the similarity in immunological sensitivity between transitional cell carcinoma and melanoma, it is of interest that similar techniques using tumor associated antigen MAGE-1- and MAGE-3-derived peptides have been successfully used for immunization in HLA-A1-positive patients with MAGE expressing melanomas, showing tumor regression in more than 30% of patients after immunization.47
The treatment of prostate cancer with peptide vaccines is relatively cutting edge, because a number of clinical trials using peptide vaccines have been completed or are currently ongoing, exploring their use in not only advanced metastatic/HRPC patients, but also in patients after prostatectomy or with biochemical relapse. For example, in 2005, Hueman et al. evaluated the E75 peptide, a HER-2/neu peptide mixed with GM-CSF in a phase I trial of high-risk patients after radical prostatectomy.48 Among 27 patients with HER-2/neu expressing prostate cancers undergoing radical prostatectomy, 17 HLA-A2+ patients received the vaccine after surgery, whereas 10 HLA-A2- patients were followed as controls. Although no clinical response was observed between the groups, the vaccine was safe and did elicit a vaccine-specific cellular and humoral immunological response. Another group tested the peptide PSA:154-163 (155 L) epitope emulsified with the adjuvant Montanide ISA-51, as a vaccination strategy for the treatment of HLA-A2+ patients with rising PSA after radical prostatectomy.49 Unfortunately, the primary end-point of the study was not met, because none of the patients showed immunoreactivity against the peptide using IFN-γ ELISPOT assays on peripheral blood mononuclear cells. In addition, no change in PSA was noted. Of note, however, several patients were found to have peptide-reactive CD8 T cell lymphocytes. In HRPC, Noguchi et al. evaluated 13 HLA-A24 positive patients who were pre-screened against a panel of 30 peptides.50 Patients were then given only their reactive peptides along with low dose estramustine phosphate. Both cellular and humoral immune responses were observed with peptide-specific CTL responses and peptide-specific IgG, along with a PSA decline in 91% of patients. This concept of using a peptide vaccine in conjunction with chemotherapy has shown some promise in pancreatic cancer.51 Miyazawa et al. used a VEGFR2 epitope peptide vaccine along with gemcitabine (which has immunomodulatory activity) in patients with metastatic, unresectable pancreatic cancer.52 This combination was found to be tolerable and resulted in CTL reactivity to the peptide in two-thirds of patients.
The power and potential of these peptide-based strategies is the ability to tailor treatments to each individual patient. Prescreening patients against a library of peptides allows the selection of only the most immunoreactive peptides, which can then be given to a patient with a particular HLA haplotype.
Limitations of existing immunotherapy
Despite preselection of peptides, which are highly immunogenic, insufficient immunostimulation is still a major problem to overcome. Many vaccines have been combined with adjuvant immune boosters to help overcome this problem to elicit a stronger immune response including Freund's adjuvant, IL-2, IL-12, GM-CSF, BCG and heat shock proteins (HSP). As discussed earlier, Provenge included GM-CSF, whereas PROSTVAC-VF included a triad of costimulatory molecules. GM-CSF, for example, has been shown to be effective to enhance peptide-specific immune reactions by amplification of peptide-presenting DC.53 Other strategies exist to enhance peptide immunogenicity, including alteration of amino acids either at or around major histocompatibility complex (MHC)-binding residues, which can increase their MHC binding affinity and T cell stimulation.54
In addition to low immunogenicity of the peptide/vaccine, there is evidence that native DC in cancer patients appear to be crippled or are targeted by tumors to evade host immune surveillance; not only do they have an impaired capacity to capture and present soluble antigens, but the also lack the costimulatory molecules necessary for T cell activation.55 Some tumors can also actively produce immunosuppressive cytokines, such as vascular endothelial growth factor, which may suppress DC activity (i.e. “suppressed phenotype”) and their ability to stimulate T cells.56 This is but one of the many mechanisms of tumor escape from immune surveillance in cancer.57,58 A full discussion of tumor immunity is beyond the scope of this article.
Novel combination therapies to boost the immune response
Emerging data suggest that primary ablative therapies can themselves induce a strong immune response. Cryoablation, radiofrequency ablation (RFA), laser interstitial therapy and high intensity focused ultrasound (HIFU), as well as ionizing radiosurgery, have all shown pro-immunological effects related to antigen shedding and the inflammatory response. Most ablative techniques primarily effect tissue destruction as a result of thermal and mechanical effects: cryoablation causes coagulative necrosis by inducing tissue temperatures to –20°C or less, whereas RFA and HIFU cause cell death by heat. The direct effects include tumor ablation/cell death and release of antigens (antigen shedding). Necrotic cells undergoing apoptosis after ablative therapy or radiotherapy become a perpetual source of tumor antigens, which can be highly immunostimulatory. HIFU also causes mechanical tumor lysis releasing additional tumor debris containing a myriad of tumor antigens. Many of the antigens that were intracellular and inaccessible to APC become “unmasked” and thereby exposed to the immune system after ablation.59 Antigen shedding appears to occur primarily as a result of disruption of cell membranes, allowing antigen leakage and heat shock induced protein denaturation. This finding was elegantly illustrated by electron microscopy after patients who were treated by HIFU under radical prostatectomy.60
Although cytokines, targeted therapy and cancer vaccines have all been used in advanced metastatic disease, ablative modalities have only been applied to localized disease. Therefore, the efficacy of combined ablative therapy and systemic immunotherapy is largely unknown, both in high-risk localized disease or the metastatic setting.
Secondary immunological effects
Although antigen shedding from ablated tumor cells can elicit systemic antitumor immunity, just outside the ablation zone or “kill zone”, non-lethal cellular injury occurs that results in a significant inflammatory response, DC activation and transcriptional upregulation of “danger signal” genes (e.g. HSP). A downstream effect of activation of “danger signal” activation is the promotion of DC maturation.61 As aforementioned, as “gatekeepers” of the immune system, DC are involved in both CD4+ and CD8+ T cell activation and T cell mediated immune responses. After appropriate costimulation events (e.g. CD40-CD40L, CD80-CD28), IL-12 production by DC, and IL-2 and interferon gamma production by CD4+ cells ensue resulting in cytolytic T cell activation.62 Both radiation and HIFU have been shown to enhance systemic antitumor immunity in situ by DC recruitment, maturation and activation inside HIFU-treated tissue.63 Hu et al. have studied the release of “endogenous danger signals” from HIFU treated tumor cells.64,65 After in vitro HIFU treatment of mouse tumor cells (MC-38), release of ATP and hsp60 from the damaged cells was detected. Exposure of HIFU treated tumor cell supernatants resulted in stimulation of APC. In addition to release of DC stimulatory factors, HIFU treated tissue can also promote DC infiltration and migration. After HIFU treatment followed by radical mastectomy, Xu et al. observed not only increased numbers of infiltrating APC (DC and macrophages), but also increased expression of costimulatory molecules CD80 and CD86 compared with control patients.66
Inflammation and danger signaling
After ablative therapy, a generalized, systemic acute phase response occurs, resulting in a non-specific boosting or ramping up of the immune response; C-reactive protein, IL-6, IL-10 and tumor necrosis factor-α (TNF-α) become elevated and immune cell recruitment ensues. Such an inflammatory cytokine cascade results in activation of monocytes and macrophages.67 Li et al. prospectively collected serum on patients undergoing prostate HIFU and cryoablation to determine levels of pro-inflammatory proteins TNF-α, IL-6 and IL-10, as well as CRP and serum amyloid A.68 Significant elevations were found for all proteins along a predictable time-course.
In addition to a generalized acute phase response, a local inflammatory response also occurs in the areas of sublethal tissue injury, just beyond the ablation zone (i.e. the border zone). An intense inflammatory infiltrate is attracted to this area, resulting in an influx of macrophages, neutrophils and lymphocytes. In addition, an increase in the Th1 to Th2 ratio has been observed favoring a specific antitumor response. Within the border zone, broad transcriptional upregulation also occurs, including growth factor and chemokine receptors, cell surface antigens and adhesion molecules, apoptosis related proteins, cell cycle regulators, HSP, and various oncogenes and tumor suppressors, enhancing tumor immunogenicity.69
Sublethal heating or damage of “out of field” tissue in areas adjacent to the target region also results in upregulation of a host of stress related genes and products including HSP.70 HSP have been found to promote strong tumor-specific T cell responses. After transrectal HIFU and subsequent radical prostatectomy, a pattern of stress protein upregulation (e.g. HSP-70) was observed, with the highest values in the border zone and the lowest values in the central necrosis region.69 Wu et al. similarly showed uniform HSP-70 upregulation in patients treated with HIFU followed by radical mastectomy.71 Thus, HSP activation occurs in the tumor milieu, which can then activate tumor infiltrating and circulating DC with induction of tumor specific cytotoxic T cells.
Both pathological cell death and non-lethal injury can lead to activation of danger signals, which include DNA damage, inflammatory and wound-healing pathways.72 Ionizing radiation has been shown to induce inflammatory-type responses and upregulate death receptors, MHC-I, costimulatory molecules such as Fas/CD95 and B7.1, various chemokines and cytokines, and NFKB mediated signals.73–76“Danger” receptors include the HSP receptor CD91, various pro-inflammatory cytokines receptors, TNF family receptors such as TNFR1 and CD40, Toll-like receptors and NKG2D receptors.77–80 Mcbride et al. noted that because radiation is given at regular intervals over a period of weeks, periodic reactivation of inflammatory and danger signaling networks occurs. These “waves” of molecular signaling create a so-called, “perpetual cascade” of cytokines, which theoretically would provide a constant source of stimulation to enhance adjuvant immunotherapeutic treatments.76
Low-intensity focused ultrasound
Increased tumor permeability: The use of low energy acoustic probes (i.e. 1 MHz) and/or variable sonication strategies using short pulsed exposure is called low-intensity focused ultrasound (LOFU), which generates low energy deposition and minimal heat. LOFU has the unique ability to alter tissue (tumor) properties through non-destructive, non-thermal mechanisms, which results in enhanced permeability of the targeted tissue.81 This occurs as a result of acoustic non-inertial cavitation, collapse, release of shock waves, microjets and radiation forces that can disrupt lipid bilayers and enhance permeability of blood vessels.70,81 In addition, intercellular gaps are widened and local hyperpolarization of the cell membrane occurs, which might increase uptake of macromolecules through endocytosis and pinocytosis.81 As a result, low frequency ultrasound promotes tissue permeability to a variety of macromolecules (e.g. drugs), which would be potentially useful in a combination therapy approach. This strategy has been applied in a few studies in non-urological cancers. For example, Bednarski et al. showed the capacity of magnetic resonance imaging guided pulsed-focused ultrasound in a rabbit model to increase target tissue permeability to 100 nm diameter liposomes containing biotin and gadolinium.82 In another study, after LOFU of subcutaneous tumor xenografts in mice, radiolabeled murine IgG mAb111I-B3 uptake by human epidermoid tumors was increased more than twofold compared with controls.83 Poff et al. showed that treatment of tumors with pulsed HIFU and the chemotherapeutic agent, bortezomib (which has been used in mRCC with limited efficacy),84 resulted in improved uptake and subsequent antitumor activity against murine squamous cell carcinoma tumors.85,86 Numerous other examples of LOFU enhanced delivery and bioavailability exist, including increased monoclonal antibody uptake, small molecules, nanoparticles, plasmid DNA and adenoviral vectors.87–91
To our knowledge, this unique property of LOFU has yet to be exploited in urological malignancy either alone or in combination with other treatments, such as targeted therapy or vaccine therapy.
The first FDA approved cancer vaccine, Provenge, has opened the floodgates to a host of novel immunotherapeutic approaches in urological malignancy including DC based vaccines and peptide vaccines. Current challenges include optimizing DC recruitment and stimulation, and generating sufficiently immunogenic peptides. Combination therapy with vaccines and danger signal-inducing treatments, such as ablative modalities, ultrasound and radiation, are promising avenues that are currently under study.