Tumor Immunology
T cell Tumor cell
Bruno Silva-Santos
[email protected] Molecular Immunology Unit
IMMUNE SURVEILLANCE OF CANCER The IS continuously patrols tumor formation
1. Lymphocytes infiltrate tumors (TIL) in mice & humans; 2. TILs (and PBLs) can kill autologous tumors in vitro; 3. Induced tumors are rejected in vivo (mice); 4. Immunodeficient mice develop higher numbers of tumors; 5. T cells (and antibodies) recognise tumor antigens.
A historical perspective of cancer immunology Background: -Pasteur (1859) and Koch (1876): “Germ theory of disease” -Metchnikoff (1883) versus von Behring (1890): Cellular vs. Humoral immunity 1891 William Coley: injection of live Streptococcus into tumors lead to regressions in 18 patients. Evolved to cell-free filtrates which he called “mixed bacterial vaccines”, but were known as “Coley’s toxins”. These became produced and marketed, and treated hundreds until 1936. 1901-1908 Jensen & Loeb: Rejection of transplanted of tumors in mice 1948 Snell & Gorer: MHC as the basis for transplant rejection - Establishment of inbred strains of mice (Strong & Little) 1957 Prehn & Main: Immune rejection of transplanted syngeneic tumors 1959 Thomas & Burnet: “Immune surveillance of cancer” theory
PRINCIPLES OF ADAPTIVE IMMUNITY TO TUMORS
Conclusions of the tumor transplantation expts: 1. All tumors are immunogenic (elicit an immune response). 2. Since it is possible to immunize against tumors, there must be tumor-specific antigens that elicit a memory response, just like against infectious agents.
1980s: CD8+ T cells isolated from melanoma lesions (TIL = tumour-infiltrating lymphocytes) can be expanded in vitro with IL-2 and lyse autologous (and HLA-matched), but not allogeneic, melanoma cells in culture.
CD8+ T cells (Cytotoxic T Lymphocytes) of the adaptive IS: - lyse tumor cells in vitro: - recognise tumor cells via peptide-MHC class I; - kill through perforins and granzymes, or Fas-ligand) - reject tumors in mice - Phenotype of CD8-deficient/ MHC-I-deficient mice; - Adoptive transfer of CD8+ T cells into nude mice.
TCRs bind MHC-peptide complexes - two large sets of very diverse molecules T cell
TCRα
TCRβ
Each TCR is specific For one MHC-peptide (pMHC) complex Peptide
TCRαβ ligand = pMHC
MHC Class I
T Cell Receptors recognise small (< 20 amino acids) peptides from digested proteins
Small peptides derived from all intracellular proteins are presented by MHC class I
All cells in our body (except erythrocytes) express MHC class I
APC
Tumor-specific versus Tumor-associated Ags (for αβ T cells) Peptide
α-chain
β2-microglobulin
Conclusions of the tumor transplantation expts: 1. All tumors are immunogenic (elicit an immune response). 2. Since it is possible to immunize against tumors, there must be tumor-specific antigens that elicit a memory response, just like against infectious agents. 3. The immunogenicity of tumors is quite variable: UV-induced > methylcholanthrene-induced > spontaneous
4. An animal that develops one type of tumor becomes immune to that tumor, but not to distinct tumors (specific response).
Mechanism of anti-tumor adaptive immunity?
Conclusions of the tumor transplantation expts: 5. Unfortunately, tumor immunization only works as prophylaxis, not as treatment: no erradication of pre-established tumor.
Prophylaxis works even started just 3 days before challenge; Treatment does not, even started 2 days after challenge.
Two obstacles to immune therapy A. Tumor changes (immunoediting & immune evasion): - Down-regulation of MHC class I expression. - Expression of inhibitory ligands for T cells. B. The IS changes: - Accumulation of regulatory T cells that suppress the killing lymphocytes.
INNATE IMMUNITY TO TUMORS
Anti-tumour lymphocytes
DC presentation CD4+ help
CD8+ T
Tumour Immunosurveillance
γδ T
NKT
NK
Innate immunotherapy of cancer: Independent on MHC class I presentation; Naturally reactive against self antigens.
Innate lymphocytes in tumor surveillance
1. γδ T cells 2. NK cells 3. NKT cells
Transformed cells often down-regulate expression of MHC I, which impairs MHCrestricted tumor-specific immunity and necessitates unconventional responses
WHAT’S THE EVIDENCE? 1. Location: found at tumor sites 2. Loss-of-function models: increased susceptibility with deficiency of subset 3. Receptor specificity: recognise ligands characteristic of transformed tissue 4. In vitro cytolytic activity directed at transformed cells
Natural born killers (NK) - inhibited by MHC class I (“Missing self”) - activated by NKG2D ligands (MICA and ULBP; RAE and H60)
- produce large amounts of IFNγ - lack of deficient mouse model - disappointing clinical trials in 1980s - current effor in small drug therapy
NKT cells in tumor immunology NKT cells activated with α-GalCer have been shown to promote: - the elimination of human melanoma, thymoma, carcinoma and sarcoma cells; - the rejection of (oncogene- or chemically-induced) tumors in mice (KO!). Cancer patients frequently show impaired number and function (IFNγ) of NKT
Main anti-tumor functions:
DC
- Production of Interferon-γ (and TNF-α); - Direct cytolysis of tumor cells;
IL-12
- Activation of NK cells; - Activation of DC
CD1d
IL-12R NK
(via CD40L-CD40)
TCR NKT
IFNγ TRAIL
for priming of CD8+ CTL. Lysis
TRAIL-R
IFNγ
CD8 T
FasL Fas
Lysis
Tumor cell Perforin & Granzymes
Immunotherapy using α-GalCer (-loaded DC) under clinical trials
γδ T cells in tumor immunology
Telma’s seminar at 13:00 hrs!
CANCER IMMUNOTHERAPY STRATEGIES
Immunotherapy of human cancer Most successful clinical trials: - autologous cancer cells + BCG (colon) - autologous tumor vaccines + anti-CTLA-4 mAb (skin, ovary) - autologous tumor lysates (kidney) - idiotype vaccines (B cell lymphoma). Promising approaches under development: - vaccines with allogeneic tumor cells - DC vaccines; DNA vaccines - small drug activation (in vivo) of lymphocytes (T, NK, NKT) - small drug/ antibody inhibition of suppressive T cells - adoptive transfer of activated (ex vivo) lymphocytes
Tumor vaccines (active immunotherapy) Vaccination with tumor antigens Active principle
Antigen
Cancer application
Peptides
NY-ESO-1 MAGE
ml,br,lu,pt,li ml,br,lu,pt,co
Proteins
HSP MAGE
ml,pa,co,re ml,br,lu,pt,co
DNA (plasmids)
Tyrosinase, gp100 CEA, MUC1
ml br,co,pa,ov,lu
Recombinant live Virus/ Bacteria
Tyrosinase, gp100 CEA
ml br,co,pa,ov,lu
Vaccination with cells: - Tumor cells - Dendritic cells
Tumor vaccines Active principle
Advantages
Disadvantages
Peptides
Easy to prepare; “Off the Shelf”
Limited epitopes; HLA requirement
Proteins
Multiple epitopes; “Off the Shelf”
Expensive; Processing
DNA (plasmids)
Easy to prepare; Multiple epitopes; “Off the Shelf”
Delivery method?
Recombinant Virus/ Bacteria
Multiple epitopes; Stimulates innate
Safety; Immunity to vector
DC
Multiple epitopes; Very immunogenic
Difficult to prepare; Maturation stage?
Autologous tumor cells
Multiple antigens; Patient-specific
Difficult to prepare (patient tissue!)
Allogeneic tumor cells
Multiple antigens; “Off the Shelf”
Alloantigens (rejection?)
Antibody immunotherapy ⇒ Chimeric and Humanized mAb
Name
Antigen
Indication (FDA)
Rituximab
CD20
B cell Non-Hodgkin Lymphoma
CD52
Chronic Lymphocytic Leukemia
Her2/ neu
Metastatic Breast Carcinoma
VEGF
Metastatic Colorectal Carcinoma
EGFR
Metastatic Colorectal Carcinoma
(Rituxan)
Alemtuzumab (Campath)
Trastuzumab (Herceptin)
Bevacizumab (Avastin)
Cetuximab (Erbitux)
Antibody immunotherapy ⇒ Bispecific mAb: BiTE (bispecific T cell engager) for CD3 (T cell) and CD19 (B cell) Non-Hodgkin’s lymphoma trial
(daily for 4-8 weeks) Bargou et al. 2008 Science 321: 974
Correlated with CD8+ T cell activation (CD69, CD25, HLA-DR)
Bargou et al. 2008 Science 321: 974
Antibody immunotherapy Liver
Bone Marrow
Control
Treated
Control
Brown = B cells (CD20 staining)
Treated
Blue = tumor cells (hematoxylin staining); Brown = T cells (CD3 staining)
Blinatumomab dose (mg/ m2 per day)
Objective response
CR vs PR
Up to 0.005
0% (n=12)
0% CR; 0% PR
0.015-0.030
20% (n=19)
10% CR, 10% PR
0.060
100% (n=7)
30% CR, 70% PR
No autoimmune side effects (≠ CTLA-4 blockade)
Tumor Necrosis Factor alpha (TNF-α) - Anti-proliferative and toxic to tumor cells - Reduces tumor blood flow and angiogenesis - Stimulates NK cells and macrophages
Cytokine therapies
Although a potent anti-tumor cytokine, severe side-effects (hypotension, Sepsis-like syndrome) have precluded extensive clinical usage. It is only administered locally in sarcomas, metastatic melanoma and liver carcinoma.
Other cytokines: GM-CSF IL-12 TGF-β IL-10 IFN-α IL-2
Activates and recruits DC; used as immune adjuvant Promotes IFN-γ responses Suppresses T and NK cell responses (prolif, cytokines) Suppresses T cell responses Inhibits tumor growth; activates DC Expands T and NKT cells
Interferon-α therapy Produced by many cell types, but 100-fold higher by plasmacytoid DC, in response to “danger” signals perceived by PRR (such as TLR) Approved for cancer immunotherapy in 1986
Induces cell cycle arrest (p53, Cdk inhibitors) and apoptosis (caspases 4, 8; Fas; TRAIL) of cancer (and viral-infected) cells Up-regulates MHC expression (HLA-I), potentially restoring it in cancer cells Promotes DC maturation, and NK and T cell activation Used as adjuvant in metastatic melanoma
Interleukin-2 therapy Growth factor (proliferation) for all lymphocytes (T, NKT, NK, B) Clinical studies since 1980s (Steve Rosenberg) showed objective response in 20% of melanoma and renal carcinoma (with 5-12% CR, lasting over 7-10 years) Approved for immunotherapy in 1992 (renal carcinoma) and 1998 (melanoma) Typical treatment: 2 cycles of 14 doses (every 8 hours), separated by 9-14 days High dose treatment initially associated with severe toxicity: cardiac depression, body edema - and 2% mortality due to bacterial sepsis, before gram+ antibiotics) Currently looking for pharmacologic ways to minimize toxicity; already reduced to fever, mild anemia and mild thrombocytosis
Immunotherapy of Melanoma
Melanoma is clearly immunogenic: - Sensitive to CTL killing in vitro; - Often regresses spontaneously in vivo; - Prognosis correlates with lymphocyte infiltrate.
-20% Objective Response - Increases overall survival and decreases risk of recurrence
-Autoimmunity (Good correlation of auto-Ab with prognosis*)
IL-2
-16% Objective Response
-Severe but reversible complications -Autoimmunity (**)
CTLA-4 blockade
Trmab: 14% OR (7% + 7%)
- Diarreha, dermatitis, pruritus
(Tremelimumab, Ipilimumab)
Ipilmab: 22% OR (9% + 13%)
-Colitis, dermatitis
Vaccines
Allovectin-7, Canvaxin, Melacine
-Failed to meet primary endpoints (OR, survival)
IFN-α
AI (52): 13% relapse, 4% death CTR (148): 73% relapse, 54% death
AI (21): 18.2 m survival CTR (28): 8.5 m survival
Adoptive cellular immunotherapy for human cancer
Difficulties:
- sufficient expansion in vitro; - nonmyeloablative conditioning regimen (“space”!); - engraftment, homing and persistence of transferred cells.
Dudley et al. 2002 Science 298: 850 13 Metastatic melanoma patients (refractory to standard therapies) Adoptive transfer of highly activated and expanded (in vitro) autologous (8x1010) CD8+ TILs plus high dose IL-2 Clonal expansion of MART-reactive clones
46% PR (skin, lung, liver, LN) 38% AID (31% Vitiligo, 7% Uveitis)
FURTHER READING: Kaufman H & Wolchok JD. General Principles of Tumor Immunotherapy. 2007. Springer. Mak T & Saunders ME. The immune response: basic and clinical principles. 2006. Chapter 26 (and chapter 30). Academic Press. Male T et al. Immunology. 7th edition, 2006. Chapter 22. Elsevier. Rosenberg SA. Progress in tumor immunology and immunotherapy. Nature 2001; 411: 380 Blattman J & Greenberg P. Cancer immunotherapy: a treatment for the masses. Science 2004; 305, 200 Ho WY, Blattman JN, Dossett ML, et al. Adoptive immunotherapy: engineering T cell responses as biologic weapons for tumor mass destruction. Cancer Cell 2003; 3:431-437. Van Der Bruggen P, Zhang Y, Chaux P, et al. Tumor-specific shared antigenic peptides recognized by human T cells. Immunol Rev 2002;188:51-64. Von Mehren M, Adams G, Weiner L. Monoclonal antibody therapy for cancer. Ann Rev Medicine 2003; 54, 343-369. Girardi M, Hayday AC. Regulation of cutaneous malignancy by γδ T cells. Science 2001; 294: 605 Kunzmann V & Wilhelm M. Anti-lymphoma effect of γδ T cells, Leuk Lymphoma 2005; 46: 671-680 Gomes AQ, Correia DV, Silva-Santos B. Non-classical MHC proteins as determinants of tumor immunosurveillance. EMBO Reports 2007; 8:1024-1030.
T cell co-stimulation for cancer immunotherapy Human in vitro and mouse in vivo studies have shown a role for: CD28/ B7 family Positive signals from CD28 and ICOS (ligands: B7.1/ B7.2 and B7h) Negative signals from CTLA-4 and PD-1 (B7.1/ B7.2 and B7-H1/ B7DC)
Blocking CTLA-4: - 80-100% rejection (prostate, breast, melanoma) in mice (plus GM-CSF vaccine) - Human trials on metastatic melanoma: 13-36% objective responses (CR, PR) - Important side-effects: autoimmunity (vitiligo, colitis, dermatitis, ...) (Note: CTLA-4 KO mice die of myocarditis due to lymphoproliferative disorder) TNF-R family Receptors CD40, CD27, OX40, 4-1BB (Ligands CD40L, CD70, OX40L, 4-1BBL)
γδ T lymphocytes Express a unique (lineage-specific), conserved TCR:
Innate immunity against pathogens: HIV, SIV HSV Epstein-Barr CMV Coxsackie Influenza Vaccinia
Plasmodium Eimeria M. tuberculosis Listeria Leishmania Chlamydia Salmonella
CD3+ CD4- CD8Humans: 2-10% of PBL (>80% of which express Vγ9Vδ2 TCR)
Play non-redundant roles in: - tumour surveillance - control of inflammation
γδ T cells in tumour surveillance Infiltrate tumour sites (lung, colon, LN…) Skin carcinomas
Express high levels of anti-tumoral effectors: - perforin, granzymes, Fas-L, TRAIL, IFN-γ
TCRδ-/-
WT
Prevent in vivo tumor progression in mice: Human γδ (Vγ9Vδ2) T cells lyse in vitro:
Tumors / mouse
18
TCRδ–/–
14 10 6 2 0
wt 0
5
10
15
Weeks after DMBA initiation
melanoma, pancreas & colon carcinoma cells (Kabelitz 2004; Corvaisier 2005) myeloma, lymphoma & leukaemia cell lines (Fish 2000; Kunzmann 2000)
NKT cells NKT cells are CD1d-restricted T cells that express NK markers and recognise glycolipids. They are innate-like lymphocytes tat produce cytokines very rapidly (1-2 hrs) after TCR ligation. Mice: Vα14Jα18 Humans: Vα24Jα18
as blocking agents
VEGF (colon)