immune checkpoint blockade in cancer therapy …...jim allison immune checkpoint blockade in cancer...
TRANSCRIPT
Jim Allison
Immune Checkpoint Blockadein Cancer Therapy
Nobel Prizein Physiology or Medicine
Lecture 2018
Peptide/MHC
CTLA-4CD28TCR
B7-1,2
No ProliferationAnergy?
T Cell
InhibitionActivation,Initiation
Dynamic Integration of TCR and Costimulatory Signalscirca 1996
IL-2IL-2p27kip
Cdk6Cdk4Cyclin D1Cyclin D2
RestrictedProliferation
Antigen Presenting Cell
APC
pRb
Cdk6Cdk4Cyclin D1Cyclin D2
pRb
S phase
Bcl-xL,g Bcl-xL,g
Gross, Harding,Krummel, Chambers, Brunner, Egen, Kuhns
Peptide/MHC
CTLA-4CD28TCR
B7-1,2
APC
Tumor
Proliferation
CTLA-4 BlockadeEnhances Tumor-Specific Immune Responses
Necrotic DeathVaccinesChemotherapyIrradiationHormone therapyAnti-angiogenesisAntibodies“Targeted” Therapies
Peptide/MHC
CTLA-4CD28TCR
B7-1,2
APC
Tumor
Attenuated orTerminated
Proliferation
CTLA-4 BlockadeEnhances Tumor-Specific Immune Responses
Necrotic DeathVaccinesChemotherapyIrradiationHormone therapyAnti-angiogenesisAntibodies“Targeted” Therapies
Peptide/MHC
CTLA-4CD28TCR
B7-1,2
APC
Tumor
Attenuated orTerminated
Proliferation
CTLA-4 BlockadeEnhances Tumor-Specific Immune Responses
Necrotic DeathVaccinesChemotherapyIrradiationHormone therapyAnti-angiogenesisAntibodies“Targeted” Therapies
APC
IL-2
UnrestrainedProliferation
The longest survivor on ipilimumabMay 2001, after progression on IL-2 10 years later
Ribas
Baseline and post-MDX-010 treatment CT scans of patient with metastatic melanoma (status post dendritic cell vaccine) who experienced regression of all known sites of disease. The patient continues without relapse at last reported
follow-up visit.
Lung Mass
Pleural Effusion
Pre-treatment Post-treatment
Ipilimumab in Metastatic Melanoma (pooled data from 4846 patients)
Patients at RiskIpilimumab 4846 1786 612 392 200 170 120 26 15 5 0
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
0 12 24 36 48 60 72 84 96 108 120
IpilimumabCENSORED
3-year OS Rate (95% CI): 21% (20–22%)
Prop
ortio
n Al
ive
Months
Schadendorf JCO 2015
Programmed Death 1(PD-1)
296 Patients with Metastatic Cancer1, 3, 10 mg/kg, MTD not reached
Safety: Adverse events similar to Ipilimumab, but 4% pneumonitis
Clinical Activity:Melamona (n= 94): 28% CR/PR, 6% SDNSCLC (n=76): 18% CR/PR, 7% SDRCC (n= 33): 27% CR/PR, 27% SDCRC (n=19), CRPC (n=13): No responses
Topalian ASCO, NEJM 2012
Anti-PD-1 Phase I (Nivolumab, BMS)
Where do we go from here?
Combinations
Ipi/Nivo vs. Ipi in Metastatic Melanoma
Hodi NEJM 2015
Immune checkpoint blockade FDA approvals
Melanoma – Ipilimumab, Pembrolizumab, Nivolumab, Ipilimumab + Nivolumab
Melanoma (adjuvant) – Ipilimumab, Nivolumab
Pediatric melanoma – Ipilimumab
Non-small cell lung cancer - Nivolumab, Pembrolizumab, Atezolizumab
Renal cell carcinoma – Nivolumab
Hodgkin’s lymphoma – Nivolumab, Pembrolizumab
Bladder cancer – Atezolizumab, Nivolumab, Durvalumab, Avelumab, Pembrolizumab
Head and neck cancer – Nivolumab, Pembrolizumab
Merkel cell carcinoma – Avelumab
MSI-H, dMMR – Pembrolizumab (any histology), Nivolumab (colorectal)
Gastric/gastroesophageal cancer – Pembrolizumab
Hepatocellular carcinoma - Nivolumab
• Determination of the cellular and molecular mechanisms involved in the anti-tumor effect
• Determination of the impact of other therapeutic agents on the immune system
• Combining the best standard-of-care therapies with immune checkpoint agents
• Targeting new molecules to improve efficacy
• Identification of predictive, prognostic or pharmacodynamicbiomarkers
Critical issues for further clinical developmentof immune checkpoint targeting
Anti-CTLA-4 Anti-PD-1
• Hard wired
• Targets CD28 pathway
• Works mainly during priming
• Expands clonal diversity
• Primarily effects CD4 T cells
• Can move T cells into “cold” tumors
• Responses often slow
• Adverse events relatively frequent
• Disease recurrence after response rare
• Induced resistance
• Targets TCR pathway
• Works mainly on exhausted T cells
• Does not expand clonal diversity
• Primarily effects CD8 T cells
• Does not move T cells into tumors
• Responses usually rapid
• Adverse events less frequent
• Disease recurrence after response significant
Can we identify checkpoint blockade responsive T cell populations?
CD8 T cells
Natural Killer cells
Myeloid cells
B cellsTregcells
CD4 T cells
NKT/γδ T cells
Dendritic cells
PD-L1-CD45-
PD-L1+ CD45-cells
Other CD45+
CyTOF analysis of murine TILs
(43 Parameters)
+/- checkpoint blockade
Unsupervised population
identification
Identify associations with
treatment and outcome
Spencer Wei
CD8 T cells
CD4 T cells
Dendritic Cells
NK cells
Myeloid cells
B cells
PD-L1+ CD45-
PD-L1-CD45-
MHC-I low Myeloid cells
CD8 T cells
NK cellsMyeloid cells
B cells
Tregcells
CD4 T cells
γδ/NKT T cells
PD-L1-CD45-
PD-L1+ CD45-
Other CD45+
B16BL6 Tumor
MC38 Tumor
Dendritic Cells
tSNE1
tSN
E2
Inoculate MC38 s.c. Antibody Antibody Antibody
Day 5 Day 8 Day 11Day 08-week old
C57BL/6
Harvest TILsDay 13
CyTOF analysis(43 Parameters)
Mass cytometry analysis of MC38 TILs
Wei et al Cell 2017
ControlαCTLA-4αPD-1
Treg
CD4 EffectorCD8
NKT/γδ
MC38 infiltrating T cell populations
Gated for CD45+, CD3e+ Wei et al Cell 2017
0
5
10
15
20
25
30
35
40
% o
f T c
ells
Control αCTLA-4αPD-1
CD44CXCR3TIM3LAP-TGFβGATA3BCL6CD127PD-L1ICOSCD69CD62LFoxP3IdUCD80MHC-ICD4CD8CD27CD25NK1.1TCRγδKLRG1CD86OX40BATFPD-1TBETRORγTEOMES
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
MC38 TIL
Wei et al Cell 2017
Checkpoint blockade modulates MC38 infiltrating T cell population frequencies
0
5
10
15
20
25
30
35
40
% o
f T c
ells
Control αCTLA-4αPD-1
Treg CD4 eff CD8 NKT
KLRG1+ Th1-likeICOSint,
, PD-1, TbetActivated CD44, CD62L
PD-1high
TIM3 Tbet
PD-1int
CD69 Tbet
Effect on TumorSize
- -Wei et al Cell 2017
Summary
• The therapeutic mechanisms of aCTLA-4 and aPD-1 are distinct
• These mechanisms are the same in a highly immunogenic and a poorly immunogenic tumor
• These distinct mechanisms may explain why the combination is so effective
• Specific CD4 and CD8 T cell subtypes contribute to the therapeutic effects in both therapies
• Monitoring these subtypes rather than total CD4 or CD8 cells correlates better with outcome and may be much predictive of outcome
Normal PBMC Ipilimumab (aCTLA-4)Nivolumab (aPD-1) Ipi + Nivo
Total CD45+ cells
TILS from treated melanoma patients
Are similar mechanisms involved in patients?
Wei et al Cell 2017
8 10 11 19 2 6 7 13 17 25 16 27 1 3 9 12 15 18 26 4 20 21 22 23 24 5 14
0
10
20
30
40
50
Clusters
% o
f CD
45Normal donor PBMCIpiPD-1Ipi + Nivo
T cells B cells Myeloid and DC Other
CD8 CD4 Other
Checkpoint blockade modulates the frequency of specific TIL populations in melanoma patientsExh CD8CD45RO+
CD69+PD-1+Lag3+ Th1-like
ICOS+ CD4CD45RO+
ICOS+PD-1low
CD69+
Treg
How do these cellular mechanisms interact?
Inoculate MC38 s.c. Antibody Antibody Antibody
Day 5 Day 8 Day 11Day 0C57BL/6
mice
Harvest TILs Day 13
CyTOF analysis(43 Parameters)
Mass cytometry analysis of MC38 TILs
Control
anti-C
TLA-4
anti-P
D-1
anti-C
TLA-4 + an
ti+PD-1
0
200
400
600
800
Tum
or v
olum
e (m
m3 )
*
Wei et al unpublished
Control
anti-C
TLA-4
anti-P
D-1
anti-C
TLA-4 + a
nti+PD-1
0
10
20
30
40
50
Freq
uenc
y of
tum
or in
filtra
ting
T ce
lls (%
)
PD-1++ LAG3+ TIM3+
Expansion of phenotypically exhausted CD8 T cells
Wei et al unpublished
Control
anti-C
TLA-4
anti-P
D-1
anti-C
TLA-4 + a
nti+PD-1
0
10
20
30
40
50
Freq
uenc
y of
tum
or in
filtra
ting
T ce
lls (%
)
*
Combination therapy differentially affects CD8 subsets
PD-1++ LAG3+ TIM3+
Wei et al unpublished
Control
anti-C
TLA-4
anti-P
D-1
anti-C
TLA-4 + a
nti+PD-1
0
10
20
30
40
50
Freq
uenc
y of
tum
or in
filtra
ting
T ce
lls (%
)
Control
anti-C
TLA-4
anti-P
D-1
anti-C
TLA-4 + a
nti+PD-1
0
10
20
30
40
50
Freq
uenc
y of
tum
or in
filtra
ting
T ce
lls (%
)
PD-1+ TIM3int Term. Diff.PD-1++ LAG3+ TIM3+
**
Wei et al unpublished
Combination therapy differentially affects CD8 subsets
Do phenotypically exhausted CD8 T cells havethe same function in the context of combination therapy?
0 200 400 600 8000
10
20
30
40
50
Tumor volume (mm3)
% o
f T c
ells
0 200 400 600 8000
10
20
30
40
50
Tumor volume (mm3)
% o
f T c
ells
Control and Monotherapyanti-CTLA-4 + anti+PD-1
Wei et al unpublished
0 200 400 600 8000
10
20
30
40
50
Tumor volume (mm3)
% o
f T c
ells
Control and Monotherapyanti-CTLA-4 + anti+PD-1
Do phenotypically exhausted CD8 T cells havethe same function in the context of combination therapy?
Wei et al unpublished
Effects on the CD4 effector compartment?
Wei et al unpublished
Control
anti-C
TLA-4
anti-P
D-1
anti-C
TLA-4 + a
nti+PD-1
0
10
20
30
40
50
Freq
uenc
y of
tum
or in
filtra
ting
T ce
lls (%
)
PD-1+ ICOSint TBET+
Th1-like CD4 effector
*
Expansion of Th1-like CD4 T cells following combination therapy
Wei et al unpublished
What is the role of costimulation in the regulation of T cell differentiation?
T cell differentiation is complexHow are phenotypes, lineages, and boundaries defined?
O’Shea and Paul. Science (2010)
Does negative costimulation regulate T cell differentiation?
Maximum signal under normal conditions
Attenuated signal due to CTLA-4 regulation
Maximum in the absence of CTLA-4?Effect on differentiation?
Ctla-4-/- T cells display distinct phenotypes
Lymph nodeCD3ε+ T cells
Wei et al unpublished
New T cells phenotypes arise in the absence of CTLA-4
WTHetKO
T cells gated
Wei et al unpublished
1 2 3 8 10 11 12 13 14 17 20 25 27 300
5
10
15
20
25
30
CD4 effector T cell clusters
% o
f T c
ells
WTHetKO
Specific expansion of CD4 T cell subsets
ICOS+PD-1+TBET+LAG3int
BCL6++ GATA3++RORγT+ICOS+PD-1+
ICOS+PD-1 low
LAG3+
CD62L+CD44-
EOMES+
GATA3int
LAG3int
Wei et al unpublished
What underlies the generation of these subsets?
• Not due to differences in T cell proliferation• Not due to differences in T cell activation• Not due to defects in thymic development
Do these populations represent new types of T cells?
Wei et al unpublished
Mass cytometry analysis of Ctla-4-/- and littermate controls
39 Parameter T cell panelActivation, surface, lineage markers
Lineage transcription factors
Comprehensive profiling of peripheral T cellsin the absence of CTLA-4
Wei et al unpublished
CD4 archetypes reside in Ctla-4-/- specific regions
t-SNE1
t-SN
E2
WT Het KO
CD4 archetypes
4 1 2 3 7 6 5
01234568
10
CD4 Archetypes
T ce
ll Fr
eque
ncy
(%)
Wei et al unpublished
Reconstruction of CD4 T cell differentiation paths
Transcription factor ratios identify lineage commitment events
Psuedotime(along T cell differentiation paths)
Wei et al unpublished
Potential implications
O’Shea and Paul. Science (2010)
Evidence for a ‘nuanced model’ of T cell differentiation
Role of T cell differentiation in mechanisms of immunotherapies
Wei et al. Cell (2017)
Inducible Costimulator(ICOS)
• Member of CD28/CTLA-4 superfamily
• Usually associated with Tfh or Treg
• Role in cancer(Sharma 2008) ICOS
CD28, CTLA-4
PD-l
ICOSL
B7-2B7-1
PD-L1 PD-L2
B7x
B7-H3
Identification of unusual ICOS+ Th1-like CD4 cells that arise after CTLA-4 Blockade
Clinical Studies• 2-10 fold increase in tumor and blood after Ipi• Contains tumor specific IFNg- & TNFa-producing CD4 cells • Increase associated with longer survival• Pharmacodynamic marker of Ipi activity
Mouse Studies• Essential for optimal efficacy of CTLA-4 blockade• Signaling via PI3K binding motif enhances Tbet expression• Can be targeted to enhance efficacy of CTLA-4 blockade
Sharma and Allison
0 20 40 60 800
20
40
60
80
100UntreatedB16B16-mICOSLaCTLA-4B16+aCTLA-4B16-mICOSL+aCTLA-4
P<0.0001
N = 20Days post tumor challenge
Perc
ent s
urvi
val
Engaging the ICOS pathway with agonist vaccine increases efficacy of anti-CTLA-4
WT
0 20 40 60 800
20
40
60
80
100Untreated (n=5)B16+aCTLA-4 (n=5)B16-mICOSL+aCTLA-4 (n=10)
Days post tumor challenge
Perc
ent s
urvi
val
ICOS KO
Fan et al. J Exp Med 2014
Combinations to enhance immune checkpoint targeting resulting in CURES
• Blocking multiple checkpoints(negative and positive)
• Enhancing innate immunity• Oncolytic viruses
• Local ablation• Blocking other immunosuppessive factors
• Conventional therapies• Radiation
• Vaccines, shared and individual• Genomically targeted therapies
% S
urvi
val
Time
Improving Survival with Combination Therapy
49
ControlStandard or Other Therapy
% S
urvi
val
50
Improving Survival with Combination Therapy
ControlStandard or Other TherapyImmunotherapy (anti-CTLA4)
Time
% S
urvi
val
ControlStandard or Other TherapyImmunotherapy (anti-CTLA4)Combination 51
Improving Survival with Combination Therapy
Time
Jane GrossFiona HardingMax KrummelDana Leach
Cynthia ChambersAndrea Van ElsasSergio Quezada
Karl PeggsTyler SimpsonMichael CurranDmitryi ZamarinSumit SubhudiXiaozho FanSpencer Wei
Allison Lab & Alumni
MedarexBristol-Meyers Squibb
The DocsThe Patients
Alan KormanPadmanee Sharma
CollaboratorsJen Wargo
Alexandre RuebenChristine Spencer
Dana Pe’er (MSKCC)Jacob Levine
NCIHHMICPRIT
SU2C/CRIPICI
MDACC Immunotherapy PlatformLuis Vence, Jorge Blando
Naveen SharmaColm Duffy
Stephen MokNana-Ama AnangCecele Denman
Alexandria Cogdill