Immune Checkpoint Blockade in Cancer Therapy …...Jim Allison Immune Checkpoint Blockade in Cancer...

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



Peptide/MHC

CTLA-4CD28TCR

B7-1,2

APC

Tumor

Attenuated orTerminated

Proliferation



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


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+


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+

**


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


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?


Effects on the CD4 effector compartment?


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


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


New T cells phenotypes arise in the absence of CTLA-4

WTHetKO

T cells gated


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


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?


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


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

(%)


Reconstruction of CD4 T cell differentiation paths

Transcription factor ratios identify lineage commitment events

Psuedotime(along T cell differentiation paths)


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


ControlStandard or Other TherapyImmunotherapy (anti-CTLA4)

Time

% S

urvi

val

ControlStandard or Other TherapyImmunotherapy (anti-CTLA4)Combination 51


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

Immune Checkpoint Blockade in Cancer Therapy …...Jim Allison Immune Checkpoint Blockade in Cancer...

Documents

Transcript of Immune Checkpoint Blockade in Cancer Therapy …...Jim Allison Immune Checkpoint Blockade in Cancer...