THERAPEUTIC CANCER VACCINES Cancer Immunology and Immunotherapy Center

THERAPEUTIC CANCER VACCINES Cancer Immunology and Immunotherapy Center St. Savas Cancer Hospital

Why we vaccinate against cancer? To activate T cell responses against TAAs that are of sufficient magnitude to eliminate tumor and prevent its recurrence

Cancer immunoediting reoccurs during immunotherapies tumor burden conventional therapy Increased OS Reduced tumor growth rates Elimination Equilibrium Anti-tumor response Generate IR M e m o r y vac start vac end Escape (recurrences)

Tumor cells can be recognized and destroyed by CD 8+ T cells, thus a therapeutic vaccine needs to activate T cells recognizing tumor antigens Cytotoxic granules CD 8+ tumor 1. 2. 3. 4. Fas. L Low numbers Low activation Short-lived immunity Wrong patient selection perforin Fas Poor clinical outcome

How can we change this? • Optimize vaccination methods to yield highly activated tumor specific CTLs in high numbers and of long-lasting anti-tumor activity Immune suppression • Combinatorial treatments • • Vaccinating patients with low tumor volume Tregs, MDSCs, enzymes/cytokines etc.

Optimizing therapeutic cancer vaccines tumor antigen ? Multiple CTL peptides from various tumor Ag restricted by various alleles. Helper peptides also needed adjuvant vehicle DCs Viral vectors

Why helper peptides? • CD 8+ induction phase and CD 8+ T cell memory • Effector phase of CD 8+ T cell at the tumor site

How to produce polyvalent vaccines 1. Synthetic long peptides encompassing CTL + TH epitopes 2. Mixtures of synthetic peptides representing immunogenic CTL + TH epitopes 3. Recombinant proteins

HER-2 (776 -790) is a promiscuous helper peptide that contains multiple HLA-DR-specific binding motifs Syfpeithi DRB 1*0101, DRB 1*1501, DRB 1*0404, DRB 1*0401, DRB 1*1101, DRB 1*1302, DRB 1*0701, DRB 1*0802, DRB 1*0901, DRB 4*0101, DRB 5*0101 Syfpeithi A 2. 1 776 GVGSPYVSRLLGICL 790 A 11, A 3 A 24 Functional assays DR 1(High), DR 3(Low), DR 4(Intermediate-High), DR 7(High), DR 8(High), DR 15(High), DR 13(High), DR 51(Low), DR 52(Low), DR 53(High) Sotiriadou NN et al. 2001 Brit. J Cancer 85(10): 1527 Salazar, LG et al. 2003 Clin Cancer Res 9: 5559 http: //www. syfpeithi. de

HER-2(776 -790) is immunogenic in vivo and has antitumor activity (Voutsas I et al, IJC, 2007, 121: 2031; Gritzapis AD et al, Cancer Res. 2006, 66: 5452; Gritzapis AD et al, Cancer Res, 2010, 70: 2686) Ii-key/HER-2(776 -790) hybrid (AE 37) is more immunogenic than native HER-2(776 -790) (Voutsas I et al, IJC, 2007, 121: 2031; Gillogly ME et al, CII, 2004, 53: 490 ) ELISPOT-IFNγ APC+HER-2(776 -790)+a. DR HER-2(435 -443) 20000 16000 12000 2 ) 250 CD 8+ + HER-2 CD 4+ 200 150 100 50 8000 0 4000 0 25 Cytotoxicity CD 8+ + Ii-key/HER-2 CD 4+ 60 50 40 CD 8+ + HER-2 CD 4+ 30 20 10 CD 8+ 20: 1 10: 1 E: T ratio 250 CD 8+ + Ii-key/HER-2 CD 4+ Tumor size (mm 2) 70 % specific 51 Cr release 50 75 100 125 Days 0 0 TUBO. A 2 HER-2(435 -443) + HER-2(776 -790) CD 8+ APC+HER-2(776 -790) Proliferation cpm A 2. 1/DR 1/neu. T+ Tumor size (mm 200 180 160 140 120 100 80 60 40 20 0 Tumor size (mm 2) no IFNγ spots (5 x 105 CD 4+ cells) Activates CD 4+ T cells in DR 4 -transgenic animals 5: 1 Days HER-2(435 -443) + HER-2(776 -790) 200 rechallenge TUBO. A 2 150 100 50 0 0 25 50 75 100 125 150 175 200 Days

Vaccinating with Ii-key/HER-2(776 -790) (AE 37) HER-2/neu+ prostate cancer patients. A phase I study

Vaccination schedule (Perez SA et al. Clin Cancer Res 2010, 16(13): 3495) Vac(id): 0. 5 ml HER-2(776 -790) (500 μg)+GM-CSF (60 μg) x 2 vac DTH Pre-DTH 0 1 2 3 Vac 2 + BD Vac 3 + BD Vac 4 + BD → 2 d (DR) Vac 1 + BD BD: blood drawing DR: dermal reaction → 2 d (DR) 4 5 6 Vac 6 + BD Post-DTH → 2 d (DR) Vac 5 + BD 12 Long term months

100 IFNγ specific spots/106 PBMC patients with response (%) Durable immunity induced by the Ii-key/HER-2(776 -790) vaccine 80 60 40 20 HER-2(776 -790)

Immunological responses and toxicity T cell subset responses Toxicity

Correlation between in vitro and in vivo immunological responses in prostate cancer patients immunized with Ii-key/HER-2(776 -790) No of responses 25 23 23 23 20 19 15 10 5 0 ICS CD 4+γ+ or CD 8+γ+ DTH Assay γ+ ICS vs DTH

Biological and immunological parameters pre vac, post vac, and at long term

Other trials using AE 37 1. 2. Phase I, NN, disease-free breast cancer patients (completed standard therapies) (Holmes JP et al. JCO 2008, 26: 3426) 2. Phase II, randomized, single-blinded, NP, NN (at high risk for recurrence), disease free breast cancer patients (Peoples GE) 3. 4. 5. Phase II, randomized, single-blinded, non-metastatic, castrate resistant prostate cancer patients in combination with anti-androgen treatment (preparing)

How to produce polyvalent vaccines 1. Synthetic long peptides encompassing CTL + TH epitopes 2. Mixtures of synthetic peptides representing immunogenic CTL + TH epitopes 3. Recombinant proteins

Vaccination against HPV-16 oncoproteins for vulvar intraepithelial neoplasia (Dutch Cancer Society, Leiden University) (Kenter GG et al, N Engl J Med 2008 361: 1838) Vaccine composition HPV-E 6 1 32 19 50 41 65 55 80 71 95 85 109 91 HPV-E 7 1 35 120 109 22 56 127 43 77 64 Adjuvant: Montanide ISA-51 140 98 158

Vaccine schedule vaccinations q 3 w x 3 -4 Each vaccine consisted of: 3. 9 mg of peptide (total 0. 3 mg x 13) in 2. 8 ml PBS Clinical Results at 24 mo after vaccination Complete responses: 9/30 Partial responses: 6/30 Kenter GG et al, N Engl J Med 2008; 361: 1838

Correlation between immunologic and clinical responses (Kenter GG et al, N Engl J Med 2009 361; 19)

Multi-peptide vaccination with IMA 901 in advanced RCC (Immatics Biotech, Gmb. H) (Walter S et al. Nat. Med. 2012; 18: 1254) A therapeutic cancer vaccine consisting of 10 synthetic tumor-associated CTL peptides, identified by: • isolating HLA-peptide complexes from primary RCC specimens • determining their sequences by MS • selection based on over-expression of their encoding genes • immunogenicity tests Plus a TH (Pan DR class II) peptide

Study design Patients with Advanced RCC (n=68) Randomization 1 1 cyclo+IMA 901+GM-CSF (i. d) placebo+IMA 901+GM-CSF (i. d) 17 vaccinations over 9 months

Immune response and clinical outcome (Walter S et al. Nat. Med. 2012; 18: 1254) Cyclo+IMA-treated patients total patients Cyclo+IMA

Phase III study – IMPRINT (Rini B et al. ASCO 2011) IMA 901 Multi-Peptide vaccine Randomized INTernational Advanced/Metastatic RCC STRATIFICATION • Risk (low vs intermediate) • Region (European vs US) • Nephrectomy (yes vs no) Sunitinib 1 st cycle Randomization 3 2 Cyclo + IMA 901 + sunitinib Sunitinib Cyclo: 300 ng/m 2 IMA 901: 10 vaccinations over 4 months Sunitinib: 5 cycles PRIMARY ENDPOINT • OS Recruiting SECONDARY ENDPOINTS • PFS-Safety • Immunomonitoring 330 patients End of enrolment: 2 nd half 2012

How to produce polyvalent vaccines 1. Synthetic long peptides encompassing CTL + TH epitopes 2. Mixtures of synthetic peptides representing immunogenic CTL + TH epitopes 3. Recombinant proteins soluble Pulsed DCs Recombinant viruses

MAGE-A 3 plus AS 15 immunostimulant (TLR-L) in patients with NSCLC (Pujol J-L et al. JCO 2012; 30(15): 7013 (suppl. May 20); ASCO Annual Meeting Proceedings) Double-blind n=122 NSCLC • stage IB (T 2 N 0) and stage II (T 1 -2 N 1), T 3 N 0) • MAGE-A 3 by q. PCR • Complete resection • Recovered (DS 0 -1) Stratified • Stage IB vs II • Squamous vs non-squamous • LN sampling vs dissection MAGE-A 3 Administration 300 μg im. • Induction q 3 wx 5 • Maintenance q 3 wx 8 • Total 27 months R 2 1 Placebo Same schedule

Total population HR=0. 73 one-sided logrank test p=0. 093 Impact of predictive gene signature on disease-free survival Gene signature negative HR=0. 78 one-sided logrank test p=0. 387 Gene signature positive HR=0. 57 one-sided logrank test p=0. 099 DFS: Interval from the date of surgical resection to the date of recurrence OR death, irrespective of cause of death HR: Hazard ratio calculated by Cox analysis

Active immunization toward the MAGE-A 3 antigen in patients with metastatic melanoma (Kruit W et al. ASCO 2011, JCO 2011; 29(15): 8535 (May 20 Suppl. )) Open-labeled/non-controlled MAGE-A 3+, unresectable stage III-IV M 1 a cutaneous Melanoma R n=36 MAGE-A 3 administration 300 mg im q 3 wx 5→q 3 mx 24 max + TLR-4 L =AS 02 B CR PR IR Med OS n=0 n=1 21% 19 months + TLR-4 L + TLR-9 L =AS 15 n=3 n=1 69% 33 months

Efficacy of MAGE-A 3 in metastatic melanoma is associated to immune microenvironment GS(+) (n=21): 10. 3 months GS(-) (n=14): 2. 3 months DFS HR=0. 31 3 6 9 12 15 18 Time (months) 21 24 27 30

Phase III study – MAGRIT (Vansteenkiste J. et al. ASCO 2007) MAGE-A 3 as Adjuvant Non-Small Cell Lun. G Cance. R Immuno. Therapy 2, 270 patients double-blind, randomized trial Resectable NSCLC Surgery Pathological stage IB, IIIA No chemo Up to 4 cycles of platinum-based chemo Randomization MAGE-A 3 + AS 15 placebo Primary endpoint: Disease free survival Secondary endpoint: Validation of gene signature predictive of response to therapy

Phase III, double blind randomized trial– DERMA (Kirkwood JM et al. J Clin. Oncol. ASCO Proceedings 2011, 29: Abstr. TPS 232) Resectable regionally advanced melanoma Stratification • Stages (IIIB, IIIC) • TNM nodal category (N 1 -N 3) • Primary category (Tx-0, T 1 -2, or T 3 or T 4) • Prior treatment (IFN, anti-CTLA 4) Surgery Randomization 2 1 1300 patients MAGE-A 3+AS 15 Placebo PRIMARY ENDPOINT • DFS SECONDARY ENDPOINT • OS, safety, health-related quality of life, validation of gene signature associated with MAGE-A 3+AS 15 benefit Schedule: q 3 w x 5 + q 12 w x 8 Recruiting

2. rec. PAP/GM-CSF pulsed on mature DCs Sipuleucel-T (Provenge®) (Dendreon) Leukapheresis In vivo T cell activation Monocytes + PAP/GM-CSF 40 hr DC infuse CTL attack tumor cells CD 4+ CD 8+ Randomized Phase 3 IMPACT Trial (Kantoff P et al, N Engl J Med 2010 363: 411) Sipuleucel-T Q 2 weeks x 3 Asymptomatic or Minimally Symptomatic Metastatic Castrate Resistant Prostate Cancer (N=512) 2: 1 Placebo 2 weeks x 3 Primary endpoint: Overall Survival Secondary endpoint: Time to Objective Disease Progression Q S U R V I V A L

IMPACT Overall Survival 100 P = 0. 032 (Cox model) HR = 0. 775 [95% CI: 0. 614, 0. 979] Median Survival Benefit = 4. 1 Mos. Percent Survival 75 Sipuleucel-T (n = 341) Median Survival: 25. 8 Mos. 50 31. 7% 25 Placebo (n = 171) Median Survival: 21. 7 Mos. 23. 1% 0 0 6 12 18 24 30 36 Survival (months) 42 48 54 60 66

Open label, randomized phase II study: sequencing of sipuleucel-T and ADT on immune responses in patients with prostatic cancer (Antonarakis S et al. ASCO 2011, 29: Abstr. TPS 189) Prostate Ca patients with rising PSA post-primary therapy, non-metastatic STRATIFICATION • PSADT (≤ 3 mo vs >3 mo) • Type of primary therapy (RP or XRT vs RP+XRT) Randomization n=30 Sipuleucel q 2 w x 3 2 wks ADT n=30 ADT 3 mo Sipuleucel PRIMARY ENDPOINT • immune responses to PA 2024 and PAP during Sipuleucel-T and after at wks 2, 6, 12 and months 6, 9, 12, 15, 18 • PSA/Testosterone: every 3 months until month 18 Accrual Dec. 2011 SECONDARY ENDPOINT • Safety • Time of PSA recurrence • Change in PSADT

Randomized, open label phase III study: sipuleucel-T plus ADT vs ADT in patients with prostate cancer (Fizazi K et al. J Clin. Oncol. ASCO 2011, 29: Abstr. TPS 188) Metastatic, androgen-dependent (hormone sensitive) prostate cancer ADT until castration-level testosterone Randomization 1 1 Sipuleucel-T + ADT n=842 ADT only n=842 PRIMARY ENDPOINT • overall survival Accrual Dec. 2011 SECONDARY ENDPOINTS • Safety • Time to castration resistance • Chemotherapy free survival

3. Recombinant proteins encoded by poxviruses Pox viruses 1. Replicate within the cytoplasm of infected cells. Do not insert their genetic sequences in the host genome 2. Infect DCs and tumor cells 3. Large genome (can accommodate large transgenes) 4. Engineered recombinants are stable TRIad TRI The PSA-TRICOM vaccine: PSA LFA-3 ICAM-1 B 7 -1 COstimulatory Molecules CO

PROSTVAC: PSA-TRICOM (Therion Biologics Corporation)

PROSTVAC: PSA-TRICOM (Therion Biologics Corporation)

Poxviral-based PSA targeted immunotherapy in metastatic castration resistant prostate cancer Phase II randomized, controlled, double blind study (NCI study) (Kantoff P et al, J Clin Oncol 2010, 28: 1094) Heterologous prime-boost regimen prime Vaccinia-PSA-Tricom boosts Fowlpox-PSA-Tricom months 0 1/2 1 2 3 4 5 GM-CSF Placebo: empty vector On PSA-TRICOM On placebo median OS (mo) median OS benefit (mo) 25. 1 8. 5 p=0. 006 16. 6

Immunologic and prognostic factors associated with overall survival employing a poxviral-based PSA vaccine in metastatic castrate-resistant prostate cancer (Gulley JL et al, CII 2010, 59: 663) 40 35 37. 3 OS (months) 30 25 26. 6 p=0. 035 20 15 10 14. 6 17. 4 20. 9 12. 3 5 0 All patients Patients with Halabi <18 mo Patients with Halabi >18 mo Halabi nomogram: LDH, PSA, alkaline phosphatase, hemoglobin, Gleacon sum , ECOG performance status, visceral disease Median survival predicted by Halabi Actual median OS

Randomized phase II study: flutamide with or without PSA-TRICOM in patients with prostate cancer (NCI) (Bilusic M et al. ASCO 2011) non-metastatic, castrate resistant prostate cancer patients with rising PSA STRATIFICATION • PSADT Randomization n=32 Flutamide + PSA-TRICOM n=32 Flutamide PRIMARY ENDPOINT • TTP (biochemical recurrence, metastatic disease)

PSA-TRICOM combined with Docetaxel and Prednisone in treating patients with metastatic castrate resistant prostate cancer (Phase II/ NCI) Schedule: Course-1 Vaccinia PSA-TRICOM Days: 1 Course-2 Fowlpox PSA-TRICOM 15 29 43 Docetaxel q 21 d Prednisone q 1 dx 2 85 - 59 Randomization Arm I: PSA-TRICOM + chemo x 12 cycles 106 (1) - (21) Arm II: chemo Objectives: Primary: OS Secondary: - Time to radiographic progression - PSA response rates - Immune responses Recruiting

PSA-TRICOM (Bavarian Nordic) Randomized double-blind Phase III Study PROSTVAC-VF Tricom + GM Asymptomatic or Minimally Symptomatic Metastatic Castrate Resistant Prostate Cancer (N=1600) PROSTVAC-VF Tricom + placebo Empty vector + placebo Primary endpoint: Overall Survival Secondary endpoint: Progression Free Survival Not recruiting yet P R O G R E S S I O N Treated at physician discretion S U R V I V A L

Optimizing the vaccine formulation for developing effective therapeutic vaccines PGE 2 ectonucleotidases PD-L 1 COX 2 i. NOS tumor antigen TMV adjuvant vehicle Adenosine gangliosides Polyvalent antigens TLR-L ? DCs Poxvirus ROS MICA/B TGF-β IL-10 IDO Arginase

Combinatorial treatments Vaccine + something chemotherapy - PSA-TRICOM + Docetaxel/Prednisone - MAGRIT (MAGE-A 3 + Platinum) - IMPRINT (IMA 901 + cyclo) - Autologous Id-KLH+GM-CSF (Biovax. ID /Biovest International)+chemotherapy in Follicular lymphoma (Phase III) (Schuster SJ et al. JCO 2011, 29: 2748) - Temozolomide (TMZ) + telomerase helper peptide (phase I/II / stage IV melanoma) (Norwegian Cancer Society) (Kyte JA et al. Clin. Cancer Res. 2011, 17: 4568) -TG-4010/MVA-MUC 1 -IL 2 (Transgene/Novartis) + chemo in NSCLC (stage IV) (Phase III) - CDX-110/EGFRv. III-KLH (Celldex Therapeutics)+ TMZ in glioblastoma (phase III) - Stimuvax BLP 25 (MUC 1+TLR-4 L in liposome) (Oncothyreon/Merck) + cyclo in NSCLC after chemoradiotherapy (Phase III)

Combinatorial treatments Vaccine + something Biological therapy tyrosine kinase inhibitors • Sunitinib + IMA 901 in RCC (IMPRINT/Phase III) (Immatics) • DCs+Au. Tu m. RNA+CD 40 L (AGS-003) + sunitinib in RCC (Phase III) (Argos Therapeutics)

Combinatorial treatments Vaccine + something hormonal therapy sipuleucel T increasing PSA level clinically localized disease • Primary therapy (radiotherapy and/or surgery) • androgen ablation metastatic disease (hormone sensitive) Androgen ablation biochemical recurrence castration-resistant disease (non-metastatic) PSA-TRICOM • monitoring • androgen ablation Secondary androgen ablation

Cancer vaccines must induce cellular immune responses before they can affect tumor burden or patient survival Immunotherapy Proper activation of the immune system Robust immunologic antitumor response Clinical efficacy Patients to be vaccinated should be sufficiently healthy for a sufficiently long time to benefit Vaccinate at earlier disease stage or in the adjuvant setting/metastatic indolent disease

Vaccinate at earlier disease stage or in the adjuvant setting/metastatic indolent disease • Phase II PSA-TRICOM study • Ongoing Phase III PSA-TRICOM study • Phase II MAGE-A 3 study in NSCLC • Phase III MAGE-A 3 MAGRIT • Phase III MAGE-A 3 DERMA • Phase III Biovax. ID study with follicular lymphoma patients • Phase III Oncophage study (autologous HSP gp 96 preparations) patients with stage I and stage II RCC → trend for increased RFS • Phase III Onco. VAX study (autologous irradiated tumor cells+BCG) patients with stage II but not stage III colorectal cancer → significant increase of RFS • Phase II E 75 study (HER-2 (369 -377)) breast cancer patients with less aggressive tumors (HER-1+, 2+, low grade) had significant increased RFS

Conclusions • Improvements in vaccine formulations • Combinatorial treatments • Vaccinating patients with low tumor burden (i. e. early disease, adjuvant setting) will improve clinical efficacy • Prognostic and predictive biomarkers are needed • Intermediate biomarkers by early disease

Acknowledgements CIIC Dr. SA Perez Dr. M Papamichail Dr. IF Voutsas Dr. AD Gritzapis Dr. P Tzonis E. Pappou E. Anastasopoulou Urology Clinic Dr. A Thanos Dr. S Bisias University of Turin Prof. G Forni Prof. F Cavallo Antigen Express Inc. Dr. R Humphreys Dr. E von Hofe University of Tuebingen Dr. C Gouttefangeas MD Anderson Medical Center TX, USA Dr. EA Mittendorf Dr. GE Peoples

Thank you – Questions?