**4.2 ONY-P1**

302 Prostate Cancer – Diagnostic and Therapeutic Advances

was shown to express glutathione S-transferase, mutant p53, CEA, and urokinase-type plasminogen activator (Warren & Weiner, 2000). GM-CSF – the cytokine the GVAX inventors have chosen as to further augment their vaccine – has shown some impact in PCa patients (PSA modulations) if administered as a solely therapeutic agent (Small et al., 1999; Dreicer et al., 2001; Schwaab et al., 2006). However, the use of GM-CSF might be challenged by counterregulatory immune responses that aim to reduce the expansion of cytotoxic T cells, thereby limiting antitumor activity. The use of GM-CSF for anti-cancer immunostimulation has caused some concerns as GM-CSF is associated with the presence of CD34+myeloid suppressor cells. Besides that it has been shown that GM-CSF is secreted by some carcinomas (Bronte et al., 1999) with a clinically relevant worse outcome (e. g. higher rate of recurrence) as in tumors with lesser CD34+cells which release Transforming Growth Factor (TGF) β inhibiting T-cell functions (Young, et al., 1997). This knowledge is important to determine the best use of GM-CSF and generally, low doses of GM-CSF are associated with greater stimulation of the immune response than higher doses which might create a counterproductive immune response via inducible nitric oxide synthase (iNOS) in well designed mouse data. This immunosupression could be abandoned by the specific iNOS inhibitor, L-NMMA, resulting in restored antigenspecific T-cell responsiveness in vitro (Serafini et al., 2004). Therefore, it is critical to optimize the use of GM-CSF, in order to

improve, rather than hamper, the immune response (Harzstark et al., 2009).

In a first phase I/II trial (coded G9802) a fixed total cell dose of 1.2 × 108 cells (6 × 107 per cell line) was used in hormone therapy-naïve patients with PSA recurrence following radical prostatectomy and absence of radiologic metastases. GM-CSF secretion from the clinical lots used in this trial was 150 ng/106 cells/24 h (LNCaP) and 450 ng/106 cells/24 h (PC-3). Patients were vaccinated weekly via intradermal injections for 8 weeks and resulted in one patient having a partial PSA response of 7 month duration. The injection sites were found to have invasion of inflammatory cells and APCs on histopathology. At 20 weeks after the first treatment, 16 of 21 treated patients showed a statistically significant decrease in PSA velocity (slope) compared with prevaccination PSA course (Simons et al., 2006) (Urba et al., 2008). Two further uncontrolled single-arm phase II studies included asymptomatic CRPC patients with (n = 34) or without (n = 21) metastases (G9803 trial) and only men with metastases (n = 80) (G0010 trial). The two trials have shown anti-tumor effects of prostate GVAX®, the first one (G9803 trial) demonstrating an overall survival benefit of 34.9 versus 26.2 months in the mCRPC subgroup (n = 34) (Small et al., 2007) and the other (G0010 trial), a study in which the vaccine was re-engineered to secrete a higher dose of GM-CSF, showing an OS ranging from 20.0 to 29.1 months (n = 80) depending on dosing regimen. Dose levels ranged from 1 × 108 cells q28d × 6 to as many as 5 × 108 cells prime/3 × 108 cells boost q14d × 11. Besides the differences in OS also the proportion of men who generated an antibody response to one or both cell lines increased with dose and included 10 of 23 in the low-dose up to 16 of 18 in the highest dose group (p < 0.01; Cochran-Armitage trend test) (Higano et al., 2008). A combined expanded retrospectively analyses of antibody response using the data from the three above mentioned trials indicated a significant (p < 0.05) association of alternative reading frame protein (TARP) antibody induction and median survival time (Nguyen et al., 2010). No dose-limiting or autoimmune toxicities were seen. The most common adverse events in both studies were injection-site erythema, myalgias, fatigue, malaise, and arthralgias. Based on these promising findings, two powerful sized randomized phase III studies of GVAX immunotherapy (VITAL-1 and VITAL-2) were set ONY-P1 (managed by Onyvax, Ltd, London, UK) consists of three irradiated PCa cell lines given to 26 patients with nonmetastatic CRPC intradermally (2.4 × 107 cells per injection), once a month for up to 12 month. In total 11 of the 26 patients demonstrated a prolonged decrease in their PSA-velocity (PSAV). None of the treated patients experienced any significant toxicity. Median time to disease progression was 58 weeks. PSAV-responding patients showed a titratable TH1 cytokine release profile in reply to restimulation with a vaccine lysate, while non-responders showed a mixed TH1 and TH2 response. Furthermore, immunologic profile correlated with PSAV response by artificial neural network analysis (Michael et al., 2005).

#### **4.3 LNCaP-IL2-IFNγ**

This approach uses only LNCaP cells retrovirally transduced with a N2/huIL2/huIFNγvector, resulting in IL-2 and IFNγ secretion. The two cytokines chosen for this approach have been used for immunostimulation solely and in combination in a variety of tumors (Brill et al., 2007; Dieli et al., 2007) and both substances are FDA approved single agents. Expression of tumor-associated antigens is upregulated after treatment with IFNγ via IFNγinducible genes, thereby increasing the susceptibility of tumors to MHC restricted CD8+CTL-mediated killing (Gansbacher et al., 1990a; Shankanan et al., 2001; Propper et al., 2003; Dunn et al., 2005). IL-2 is a well-known T cell growth factor, which is traditionally implicated in the agonistic stimulation of immune responses (Gansbacher et al., 1990b; Rosenthal et al., 1994) and FDA approved for systemic application against metastatic renal cell cancer. IL-2 is the only cytokine to date not been detected to be produced by any cancer. LNCaP cells, as mentioned above (see GVAX) express some relevant TAAs but in contrast to other PCa cell lines used as cancer vaccines, do not express transforming growth factor (TGF)-β. After detailed investigation of the safety profile with special attention to induction of autoimmunity (n = 3 patients at a dose level of 7,5 × 106 cells) (Brill et al., 2007), further patients (n = 27) were scheduled to receive four intradermal vaccine injections (dose level of

Entering a New Era – Prostate Cancer

Immuno-Therapy After the FDA Approval for Sipuleucel-T 305

and cytolysis. Therefore, combining the elimination of Treg-cells – achieved in specific (via mAb) or unspecific (via low dose chemotherapy) manor – with therapeutic immunotherapy approaches seems promising. When comparing DC vaccination in patients with renal cell carcinoma (RCC) with/without previous elimination of Treg-cells by an IL-2 diphtheria toxin conjugate (Denileukin diftitox; ONTAK®), a higher immune response rate was achieved in the Denileukin diftitox group (Dannull et al, 2005). Low, i.e. not tumoricidal doses of chemotherapy can also synergize with vaccination: It could be shown that low dose cyclophosphamide selectively diminishes quantity and function of Treg-cells with little effect on other lymphocyte subpopulations (Ghiringhelli et al., 2007). Besides low dose chemotherapy it has been noticed in individuals with RCC that the multi-kinase inhibitor sunitinib shows depletion of Treg-cells (Finke et al, 2008; Kusmartsev & Vieweg, 2009). Besides Treg-cells (CD4+CD25highFoxP3+T-cell) other subsets of immune cells with suppressive function are focused in preclinical studies (e.g. CD4+NKT-cells, MDSCs, tumor

*PD-1* is another co-inhibitory receptor molecule expressed on activated T-lymphocytes and functioning as an immune checkpoint. When PD-1 is bound by its ligand, T-cell proliferation and activation are inhibited, resulting in inhibition of anti-tumor immune responses (Blank & Mackensen, 2007). Expression of the PD-1 ligand has been described on a variety of human tumor cells including PCa, leading to decreased tumor-specific immunogenicity (Ebelt et al., 2009). In addition, expression of the PD-1 ligand correlates with a poorer prognosis in some human malignancies. MDX-1106, a fully human anti-PD-1 blocking antibody, is the first agent in its class to reach human clinic. Results of a phase I trial using MDX-1106 in patients with refractory metastatic solid tumors (including metastatic CRPC) were recently published and shows signs of immunological impact and clinical response – but not jet in the included PCa patients (Brahmer et al., 2010). Common side effects of this agent – but also seen in trials blocking CTLA-4 and after depleting Treg-cells – include

Other targets but to date only in preclinical research are TH17 cells and B7-inhibitors. For TH17cells it is known that TGF-β together with IL-6 in addition to IL-1-β, IL-21 or IL-23 promotes development of TH17 cells (OConnor et al., 2010). The role of that TH-subtype in cancer vaccine settings is not clear yet. They have a well defined function in autoimmune or autoinflammatory-associate pathology as they at least in some aspects (e.g. production of IL-10) act like Treg-cells. The frequency of IL-17 producing T-cells might correlate with clinical response to therapeutic vaccination in CRPC patients (Derhovanessian et al., 2009). B7 is a co-signal important in T-cell development (see Prostvac-VF) and B7-inhibitors are constitutively expressed in numerous types of human tumors and had been shown to support evasion of tumor immunity by promoting apoptosis of activated effector T-cells, inhibiting IL-2 production and tumor resistance to T-cell mediated lysis. B7-inhibtors can be blocked by new drugs tested in preclinical models resulting in enhanced cytotoxic T-cell

Immunostimulatory trials in various cancer entities has often been judged as limited in the impact but have always been seen as at least non toxic and so not harming the patients. AEs seen in vaccine trials have in most cases be limited to local injection side reactions. Vaccines which use a cytokine in addition reported flu-like symptoms but not above grade 2. If

associated Macrophages, and CD4+TH2-cells) (Palena & Schlom, 2010).

subclinical hypothyroidism, colitis and autoimmune arthritis.

responses against TAAs (Palena & Schlom, 2010).

**6. Points to consider: Safety, regulatory and ethical issues** 

1.5 × 107 cells) on days 1, 15, 29, and 92. In the absence of disease progression, patients received further vaccinations every three months – resulting in a total amount of more or less 100 million cells/year – as compared to 4,600 million used in other trials (see GVAX). The primary study criteria were safety and the difference in PSA-DT, determined in the pretreatment phase (before the start of vaccination) and in the trial treatment phase (during vaccination). During vaccination there was a significant prolongation of the PSA-DT compared with the prevaccination period (from 63 to 114 days (p < 0.01; intention to treat analyses)). The median overall survival time from first vaccination was 32 months the median Halabi-predicted survival in these thirty heavily pretreated mCRPC-patients was calculated as low as 15 months. Despite such comparisons of assessed and Halabipredicted survival are widely accepted to compare patients groups and even different approaches (Lubaroff et al., 2009; Gulley et al., 2010; Kantoff et al., 2010b) these data are to be categorized somehow as speculative. ELISPOT based immune monitoring revealed Tcell stimulation in the majority of patients and artificial network analysis showed best predictive value for a response to the vaccination for survirvin-reactivity at day 36 (Brill et al., 2009).

#### **5. Approaches to overcome tumor induced immunosuppression**

Three strategies have reached clinical phases in this respect: i) Inhibition of CTLA-4, ii) depletion of Treg-cells, and blocking programmed death-1 (PD-1) checkpoint.

*CTLA-4* is a T-cell surface glycoprotein that is up regulated following T-cell activation to restrain the immune response. Its main task is to prevent autoimmunity by regulating the body's immune response to self antigens. Two fully human mAbs focussed to block CTLA-4 are at this time in clinical development: ipilimumab (MDX-010; Medarex/Bristol-Myers Squibb, Princeton, NJ) and tremelimumab (CP-675,206; Pfizer, New York, NY). The later is an immunoglobulin G2 antibody and ipilimumab is an immunoglobulin G1κ antibody; both bind to CTLA-4 with affinities less than 1 nmol/L. T-cells express two counteracting receptors on their cell surface – CD28 and CTLA-4. Both bind to the same ligands or co-stimulatory molecules on the outside of APCs (B7.1 and B7.2). Binding to CD28 results in activating T-cells, while interacting with CTLA-4 inhibits T-cell stimulation. In the first human phase I trial the use of an anti–CTLA-4 antibody alone in men with PCa, 14 patients with progressive mCRPC, seven of whom had received prior chemotherapy, were given one single dose of ipilimumab. Two patients demonstrated PSA declines of > 50 % and two patients established prolongation of their PSA-DT (Small et al., 2007b). These data suggest – as these end stage patients got no other therapy besides anti-CTLA-4 mAb – that some degree of immune "autopriming" exists even in a difficult stage of disease, with an ongoing low-level presentation of antigen, perhaps from basal apoptotic rates, which can then be enhanced with CTLA-4 blockade (Harzstark et al., 2009). Blocking CTLA-4 has been shown to prolong and potentiate immune responses to vaccine in two phase I trials (one as combination with GVAX; the other as combination with Prostvac-VF) (Arlen et al., 2009). Combinations of ipilimumab with chemotherapy as if vaccine has been successful in malignant melanoma and ipilimumab was recently approved for this purpose by the FDA (Hodi et al., 2010; Robert et al., 2011).

*Treg-cells* are induced during vaccination and the expansion of Treg-cells already present in the patient are possible factors causative to impaired vaccine efficacy by suppressing CD4+ and CD8+ effector T-cells through IL-2 consumption, the secretion of suppressive cytokines,

1.5 × 107 cells) on days 1, 15, 29, and 92. In the absence of disease progression, patients received further vaccinations every three months – resulting in a total amount of more or less 100 million cells/year – as compared to 4,600 million used in other trials (see GVAX). The primary study criteria were safety and the difference in PSA-DT, determined in the pretreatment phase (before the start of vaccination) and in the trial treatment phase (during vaccination). During vaccination there was a significant prolongation of the PSA-DT compared with the prevaccination period (from 63 to 114 days (p < 0.01; intention to treat analyses)). The median overall survival time from first vaccination was 32 months the median Halabi-predicted survival in these thirty heavily pretreated mCRPC-patients was calculated as low as 15 months. Despite such comparisons of assessed and Halabipredicted survival are widely accepted to compare patients groups and even different approaches (Lubaroff et al., 2009; Gulley et al., 2010; Kantoff et al., 2010b) these data are to be categorized somehow as speculative. ELISPOT based immune monitoring revealed Tcell stimulation in the majority of patients and artificial network analysis showed best predictive value for a response to the vaccination for survirvin-reactivity at day 36 (Brill et

**5. Approaches to overcome tumor induced immunosuppression** 

depletion of Treg-cells, and blocking programmed death-1 (PD-1) checkpoint.

approved for this purpose by the FDA (Hodi et al., 2010; Robert et al., 2011).

*Treg-cells* are induced during vaccination and the expansion of Treg-cells already present in the patient are possible factors causative to impaired vaccine efficacy by suppressing CD4+ and CD8+ effector T-cells through IL-2 consumption, the secretion of suppressive cytokines,

Three strategies have reached clinical phases in this respect: i) Inhibition of CTLA-4, ii)

*CTLA-4* is a T-cell surface glycoprotein that is up regulated following T-cell activation to restrain the immune response. Its main task is to prevent autoimmunity by regulating the body's immune response to self antigens. Two fully human mAbs focussed to block CTLA-4 are at this time in clinical development: ipilimumab (MDX-010; Medarex/Bristol-Myers Squibb, Princeton, NJ) and tremelimumab (CP-675,206; Pfizer, New York, NY). The later is an immunoglobulin G2 antibody and ipilimumab is an immunoglobulin G1κ antibody; both bind to CTLA-4 with affinities less than 1 nmol/L. T-cells express two counteracting receptors on their cell surface – CD28 and CTLA-4. Both bind to the same ligands or co-stimulatory molecules on the outside of APCs (B7.1 and B7.2). Binding to CD28 results in activating T-cells, while interacting with CTLA-4 inhibits T-cell stimulation. In the first human phase I trial the use of an anti–CTLA-4 antibody alone in men with PCa, 14 patients with progressive mCRPC, seven of whom had received prior chemotherapy, were given one single dose of ipilimumab. Two patients demonstrated PSA declines of > 50 % and two patients established prolongation of their PSA-DT (Small et al., 2007b). These data suggest – as these end stage patients got no other therapy besides anti-CTLA-4 mAb – that some degree of immune "autopriming" exists even in a difficult stage of disease, with an ongoing low-level presentation of antigen, perhaps from basal apoptotic rates, which can then be enhanced with CTLA-4 blockade (Harzstark et al., 2009). Blocking CTLA-4 has been shown to prolong and potentiate immune responses to vaccine in two phase I trials (one as combination with GVAX; the other as combination with Prostvac-VF) (Arlen et al., 2009). Combinations of ipilimumab with chemotherapy as if vaccine has been successful in malignant melanoma and ipilimumab was recently

al., 2009).

and cytolysis. Therefore, combining the elimination of Treg-cells – achieved in specific (via mAb) or unspecific (via low dose chemotherapy) manor – with therapeutic immunotherapy approaches seems promising. When comparing DC vaccination in patients with renal cell carcinoma (RCC) with/without previous elimination of Treg-cells by an IL-2 diphtheria toxin conjugate (Denileukin diftitox; ONTAK®), a higher immune response rate was achieved in the Denileukin diftitox group (Dannull et al, 2005). Low, i.e. not tumoricidal doses of chemotherapy can also synergize with vaccination: It could be shown that low dose cyclophosphamide selectively diminishes quantity and function of Treg-cells with little effect on other lymphocyte subpopulations (Ghiringhelli et al., 2007). Besides low dose chemotherapy it has been noticed in individuals with RCC that the multi-kinase inhibitor sunitinib shows depletion of Treg-cells (Finke et al, 2008; Kusmartsev & Vieweg, 2009). Besides Treg-cells (CD4+CD25highFoxP3+T-cell) other subsets of immune cells with suppressive function are focused in preclinical studies (e.g. CD4+NKT-cells, MDSCs, tumor associated Macrophages, and CD4+TH2-cells) (Palena & Schlom, 2010).

*PD-1* is another co-inhibitory receptor molecule expressed on activated T-lymphocytes and functioning as an immune checkpoint. When PD-1 is bound by its ligand, T-cell proliferation and activation are inhibited, resulting in inhibition of anti-tumor immune responses (Blank & Mackensen, 2007). Expression of the PD-1 ligand has been described on a variety of human tumor cells including PCa, leading to decreased tumor-specific immunogenicity (Ebelt et al., 2009). In addition, expression of the PD-1 ligand correlates with a poorer prognosis in some human malignancies. MDX-1106, a fully human anti-PD-1 blocking antibody, is the first agent in its class to reach human clinic. Results of a phase I trial using MDX-1106 in patients with refractory metastatic solid tumors (including metastatic CRPC) were recently published and shows signs of immunological impact and clinical response – but not jet in the included PCa patients (Brahmer et al., 2010). Common side effects of this agent – but also seen in trials blocking CTLA-4 and after depleting Treg-cells – include subclinical hypothyroidism, colitis and autoimmune arthritis.

Other targets but to date only in preclinical research are TH17 cells and B7-inhibitors. For TH17cells it is known that TGF-β together with IL-6 in addition to IL-1-β, IL-21 or IL-23 promotes development of TH17 cells (OConnor et al., 2010). The role of that TH-subtype in cancer vaccine settings is not clear yet. They have a well defined function in autoimmune or autoinflammatory-associate pathology as they at least in some aspects (e.g. production of IL-10) act like Treg-cells. The frequency of IL-17 producing T-cells might correlate with clinical response to therapeutic vaccination in CRPC patients (Derhovanessian et al., 2009). B7 is a co-signal important in T-cell development (see Prostvac-VF) and B7-inhibitors are constitutively expressed in numerous types of human tumors and had been shown to support evasion of tumor immunity by promoting apoptosis of activated effector T-cells, inhibiting IL-2 production and tumor resistance to T-cell mediated lysis. B7-inhibtors can be blocked by new drugs tested in preclinical models resulting in enhanced cytotoxic T-cell responses against TAAs (Palena & Schlom, 2010).
