Dendritic Cell-Based Immunotherapy Vaccines for Melanoma and Hepatocellular Cancer

Transcription

1 Dendritic Cell-Based Immunotherapy Vaccines for Melanoma and Hepatocellular Cancer Lisa H. Butterfield, Ph.D. Assistant Professor of Medicine, Surgery and Immunology University of Pittsburgh Cancer Institute Director of Operations, UPCI Immunologic Monitoring and Cellular Products Laboratory

2 Dendritic Cells at the center of the immunological universe: 1. Sampling their environment 2. Sensing pathogens 3. Trafficking from the periphery to lymph nodes 4. Presenting antigen and shaping the adaptive immune response 5. Inhibiting unwanted responses (tolerance) and activating needed responses 6. Many different types of DC

3 Differentiation of Dendritic Cells Myeloid Precursor CD1- CD14++ CD83- B7-1- B7-2+ Class I+ Class II+ CD54+ CD40- IL-12- GM-CSF IL-4 Immature DC CD1+ CD14+ CD83+/- B7-1+ B7-2++ Class I++ Class II+++ CD54++ CD40++ IL-12- CD40L TNFa IFNa LPS CpG oligos Mature DC Ca Ionophore CD1++ Cytokine CD14- cocktails CD83+ etc. B7-1++ B Class I+++ Class II++++ CD54+++ CD40+++ IL-12+

4 MART Peptide-pulsed DC Clinical Trial: Phase I/II 7/97-4/01; Clin.Ca.Res., 3/03 5/01-4/02; J. Immunother., 9/04 PBMC MART DC GM-CSF + IL-4 x 3 PBMC: -ELISPOT -MHC Tetramer Assay -FASTIMMUNE -cytotoxicity Which correlates with clinical response? MART-1+/HLA-A2.1+ Phase I: 10 5, 10 6, 10 7 DC/injection i.v. vs. i.d. at each dose (18 pt.) Phase II: 10 7 DC/injection, i.d. (10 pt.)

5 Patient E1 (10 7 DC, i.d.) post: 6 surgeries, 32 doses radiation, 6 infusions IFNα Pretreatment +56 days +130 days Melanoma Tumor Lymphocytic Infiltrate (largely CD8+, also CD4+) Absence of Melanoma

6 M A R T 2 7 tyro s1 A F P g p Post-treatment Fold Increase/Pretreatment B1 DETERMINANT SPREADING: MART peptide pulsed DC Trials B2 B3 C1 D2 E1 * * E2 F2 F3 * G 1 G2 G 3 G4 * G5 G 6 G8 G 9 G 10 gp MART no peptide AFP Tyros. 1-9 gp MART HLA-A2.1 CLASS I DR4 CLASS II

7 Dendritic Cell-based Strategies 1. Lysate/Protein 2. Eluted/Purified Peptide 3. Transfection 4. Tumor Fusion + Endogenous processing HLA type independent MHC class I and II HLA type dependant Multiple antigens MHC class I only Endogenous processing HLA type independent Multiple antigens MHC class I and II Endogenous processing HLA type independent Multiple antigens (tumor, housekeeping) MHC class I and II

8 DENDRITIC CELL-BASED STRATEGIES 1. Lysate/Protein Endogenous processing/multiple epitopes, HLA type independent MHC class I and II (lysate is a complex, undefined mixture) 2. Eluted/Purified Peptide 3. Transfection 4. Tumor Fusion + HLA type dependant Multiple antigens (synthetic are defined), specific MHC class I/II, peptide half-life? Endogenous processing/multiple epitopes, HLA type independent, defined antigens, in MHC class I and II Endogenous processing/multiple epitopes, HLA type independent Multiple antigens (tumor, housekeeping) MHC class I and II

9

10 Genome structures of first generation, second generation and gutted / gutless / helper-dependent adenovirus vectors.

11 GENETIC ENGINEERING OF DENDRITIC CELLS MART-1 AdVMART1 MART DENDRITIC CELL CD4+ CD8+ MART-1 T Cells Cytokine production Proliferation Cytotoxicity HLA A2.1/DR4 MPREDAHFIY GYPKKGHGHS YTTAEEAAGI GILTVILGVL LLIGCWYCRR RNGYRALMDK SLHVGTQCAL TRRCPQEGFD HRDSKVSLQE KNCEPVVPNA PPAYEKLSAE QSPPPYSP

12 MART-1 AdV-Transduced Dendritic Cell Clinical Trial leukapheresis MART-1 Adenovirus DC GM-CSF + IL-4 3 vaccines, i.d. CD8+/CD4+ PBMC: -MART-1 ELISPOT -Tetramer Assay -Determinant Spread ELISPOT 3/02-3/04 (23 enrolled) 14 received all 3 vaccines 1 unevaluable (54+ mo.), 4 SD (27, 33, 36, 42 mo.) J. Immunotherapy 08

13 Net IFNg spots per million cells * A1 * net IFNg spots per million cells B6 0 0 pre d14 d28 d35 d56 pre d14 d28 d35 d56 pre d14 d28 d35 d56 d+112 pre d14 d28 d35 d56 d+112 CD8+ CD4+ CD8+ CD4+ DC vaccines B3 and B4 CD8+ ELISPOT Net IFNg spots per million cells pre d14 d28 d35 d56 pre d14 d28 d35 d56 High baseline response CD8+ CD8+ Examples of AdVMART1/DC patient CD8+ and CD4+ T cell responses to MART-1 peptides by IFNγ ELISPOT.

14 Assessment of Determinant Spreading in AdVMART1/DC Patients Net IFNg spots per million cells MART27 gp tyros 1-9 MAGE3 MART51 gp pre post pre post pre post pre post pre post A1 B1 B2 B4 B6 PD, OS = 14 mo. Decreasing responses to other antigens PD, OS = 3 mo. Decreasing responses to other antigens SD, OS = 36 mo. Increasing CD8 responses to tyrosinase PD, OS = 58+ mo. (NED) Increasing CD8 responses to gp100 UnEv, OS = 56+ mo. Increasing CD8 and CD4 responses to gp100, CD8 response to MAGE-A3

15 Immune Response, Clinical Response and Determinant Spreading Immune response to the vaccine-encoded antigens can correlate with clinical response, but the magnitude of CD8+ T cells to a specific immunizing peptide rarely defines the clinical responders. Total immune score, breadth of immunity and determinant spreading may be more critical. How to strengthen and broaden immunity: 1. Immunize with multiple antigens 2. Include CD4 helper stimulation 3. Further improve the DC (subtype, maturation) 4. Promote innate immunity (NK cells) 5. Select antigens critical to tumor function?

16 Possible mechanisms of AdV/DC improved antigen presentation and T cell activation: Full length gene transfer: 1. Provision of antigen in both MHC Class I (CD8 T cells) and MHC Class II (CD4 T cells). 2. Long-term antigen synthesis (7-14 days, vs. hours for peptides). Viral Effects on DC: 3. Modified APM for more efficient processing and presentation? 4. DC maturation (cell surface phenotype improved T cell signals, cytokine and chemokine production) what works with AdV? 5. AdV/DC and NK cell interactions?

17 Inside the DC Vaccine:AdV-transduced DC and Antigen Processing Machinery (APM) MHC class I Heavy Chain Calnexin β2m Golgi ERp57 TAP1 TAP2 Calreticulin Tapasin Proteasome Peptides Δ, MB1, Z LMP2, LMP7, LMP10 Ub Protein APM antibodies from S. Ferrone IFN-γ

18 Unst. CD86 CCR6 Cell Surface Phenotypic Markers DR CD40 CCR7 ABC CCR4 CD83 idc mdc CD80 CCR5 AdVeGFP/DC EGFP

19 Intracellular APM Proteins Unst. β2m LMP2 LMP7 Calnexin Calreticulin Proteins from the proteosome, the immuno-proteosome, TAP peptide transporters and chaperones. Delta LMP10 ERp57 Adenovirus transduction of DC increases expression of TAP1/2 peptide transporters and ERp57 peptide chaperone. MB1 Z TAP1 TAP2 Tapasin AdVeGFP/DC moi = 500 ERp57 is an oxidoreductase which regulates both the construction and destruction of peptide- MHC class I antigen complexes

20 NK/DC Interactions: Literature: Activated NK can stimulate DC NK can kill immature DC Mature DC can activate NK cells Question: AdV-transduced DC?

21 NK Cell Subsets Are Activated By Ad.DCIn an AdV Dose-Dependent Manner: CD69 Expression 400 CD69 CD56hi CD16- CD56dim CD MFI NK alone NK + idc NK + Ad.DC 100 NK + Ad.DC 500 NK + Ad.DC 1000 NK + mdc Ad.LacZ

22 NK Cells Activated by Ad.DC More Effectively Recognize K562 Than Those Activated by idc IFN-g Spots Per 10,000 NK Cells NK only NK + idc NK + Ad.DC NK + mdc Media Only K562

23 idc and Ad.DC-Induced NK Cell Activation Is Mediated by Cell-to-Cell Contact idc Ad.DC mdc CD56 dim CD16 + MFI MFI MFI CD56 hi CD16 - MFI MFI MFI NK only NK/DC NK/DC Transwell

24 How to strengthen and broaden immunity to improve clinical outcome? 1. Immunize with multiple antigens and 2. Add CD4 helper stimulation (tumor antigen) and 3. Improve the DC and 4. Promote innate immunity and 5. Promote spreading to additional tumor antigens.

25 Immunohistochemistry of AdVTMM Expression of Tyrosinase, MART-1 and MAGE-A6 in Human DC Mel526 DC only AdVTMM Single Gene AdV Tyrosinase antibody AdVTyrosinase MART-1 antibody AdVMART1 MAGE-A6 Antibody AdVMAGE6 No Antibody

26 Melanoma Cell Recognition by AdVTMM/DC Stimulated T cells Mel526 ELISPOT TMM MART TMM MAGE TMM MART TMM MAGE TMM Tyr CD8 IFNg CD8 IFNg CD4 IL2 CD4 IL2 CD4 IL2 CD4 IL2 CD4 IFNg CD4 IFNg CD4 IFNg CD4 IFNg spots per 10e5 cells Short term cultures stimulated with AdVTMM/DC recognized Mel526 (expresses all 3 antigens) better than AdV single antigen /DC.

27 Multiple, defined antigen loading strategies for DC Protein-pulse Glypican-3 Peptide-pulse MAGE-A3 AdV Transduce AdVhAFP Tumorderived AFP? How does route of uptake change the mechanism of processing and presentation? Can you protein-pulse, peptidepulse AND AdV-transduce and get all antigens (and epitopes) adequately presented? Will protein-fed promote MHC class II/CD4 immunity preferentially (due to endosome pathway)? Will AdV cytoplasmic expression promote MHC class I/CD8 immunity preferentially? Will helped CD8 be superior?

28 T = hours optical slicing AdVhAFP/DC AFP = green, Actin = red, nucleus = blue

29 T = hours DC only (actin=green or red) AdVhAFP/DC (AFP = red, actin= green) AdVLacZ/DC (-) (actin= green) GPC3/DC (GPC3 = green, actin=red) BSA/DC (-) (actin=red) AdVhAFP/GPC3DC (GPC3 = green, AFP=red)

30 AdVhAFP+GPC3 protein DC T = hours confocal GPC3 = green, AFP = red, Actin = blue

31 Superior CD8+ Cytotoxicity with AdV/DC vs. Protein/DC CD8+ Granzyme B ELISPOT K K K K K K562+0 AFP/DC DC+0 net spots per 10e5 CD AdVhAFP AFPprotein AdVhAFP/DC were able to expand AFP-specific, cytotoxic (granzyme B+) CD8+ T cells to all 4 immunodominant epitopes tested. AFP protein/dc were able to expand AFP-specific CD8+ T cells specific to 2 of the 4 immunodominant epitopes tested. The relative epitope dominance differed between the two modes of antigen presentation. AdV/DC were also superior for activation of CD8+ T cells recognizing AFP protein-pulsed DC as APC.

32 Antigen loading/processing pathway on CD4+ T cell responses CD4+ AFP responses IFNg CD4 DC+0 CD4 AFP/DC CD4 AFP/DC CD4 DC+AdVhexon AdVhAFP AFPprotein IFNg AFP protein/dc were a superior stimulus for AFP-specific CD4 T cell expansion. AdVhAFP/DC were able to expand AFP-specific, IFNg+ CD4+ T cells. As expected, the AdV/DC boosted memory and effector AdV-virus-specific T cells.

33 AdV/DC have fewer Treg than protein/dc or peptide/dc 30 Flow Treg AdVhAFP AFPprotein AFP pep 30 AdVhAFP GPC3 GPC+AdV Flow Treg analysis %+ 15 % %CD4/CD25+ %CD4/CD25hi CD4/CD25+%FoxP3 Activated Treg FoxP3 in CD4+ 0 %CD4/CD25+ %CD4/CD25hi CD4/CD25+%FoxP3 Activated Treg FoxP3 in CD4+ AdV/DC have slightly increased activated CD4+ (CD25+), but reduced Treg (CD4+/CD25hi), and the least amount of CD4+/CD25+ or CD25hi+/FoxP3+ cells. Protein-fed alone had highest Treg, and adding AdV transduction to protein reduced the Treg numbers to similar to AdVhAFP/DC. Peptide/DC are similar to protein/dc, with higher Treg.

34 Conclusions First generation cancer vaccines have been safe and immunogenic. Many trials find weak and transient CD8 T cell function and a lack of correlation between the frequency of single antigen vaccineinduced T cells and clinical outcome. Next generation vaccines can examine the impact of antigen loading, more subtle DC manipulations and multiple types of effector interactions to generate more potent anti-tumor immunity which will be more effective clinically. Consistent lack of toxicity may also allow testing in earlier stage patients and open up the opportunity for prevention studies.

35 Acknowledgements UPCI/Butterfield Lab Lazar Vujanovic, Ph.D. Jian Shi, M.D. Sarah Newton, B.A. Jason Tse Jennifer Landsberg Univ. Pittsburgh Center for Biological Imaging (S. Watkins) UPCI Immunological Monitoring Laboratory UPCI Luminex Core Facility (A. Lokshin) UPCI Flow Cytometry Facility (A. Donnenberg) (UCLAclinical trial data: with J. S. Economou and A. Ribas) University of Pittsburgh Cancer Institute, NIH RO1 CA American Heart Association, The Pittsburgh Foundation, The Hillman Foundation