Basic Concepts and Future Horizons in Cancer Immunotherapy

Drew Pardoll

Johns Hopkins

The amplitude of immune responses is determined

by a balance of positive signals (antigen+costimulation)

and negative forces (immune checkpoints)

IMMUNE CHECKPOINTS = BRAKES

Stat3, ~20 receptor/ligand, Treg, MDSC

Immune

ResponseIMMUNIZATION = ACCELERATOR

Signal 1 – Antigen

Signal 2 - Costimulation

Dendritic Cells VACCINE

Why do some

vaccines induce

tumor-specific

immunity but

rarely induce

tumor regression?

B7.1/B7.2

?

PD-L1/PD-L2

B7RP -1 ICOS

CD28

B7H3 ?

MHC/pep TCR

CTLA-4

PD-1

Signal 1

CD137L CD137

OX40L OX40

LIGHT LIGHT-R

Antigen

Presenting Cell

or Tumor Cell

T Cell

Cytokines (IL-1, IL-6, IL-10, IL-12, IL-18)

CD200R CD200

CD40 CD40L

HVEM BTLA

The Immunologic Synapse

PS/galectin9 Tim1/Tim3

LAG3

APC T cell

Activation(cytokines, lysis, prolif., migration)

B7.1 CD28

TCR Signal 1MHC-Ag

TumorTumor

PD-L1

PD-1

(-)(-)

(-)

Inhibition(anergy, exhaustion, death)

Anti-

PD-1

The PD-1 Pathway is a Master Checkpoint

within the Tumor Immune Microenvironment

Tumor

DC/M

maturation

LAG-3

CD4 T cell

Treg

Anti-

tumor CD8

killer cell

MDSC

PD-L1

PD-L1

PD-L2

A2aR

Multiple Cell Types within the Tumor Immune

Microenvironment Express PD-1 and PD-1L

PD-1

LAG-3

Tim3

BTLA

PD-1

B7-H3B7-H4

PD-L1/2PD-L1

IDO

IDOTTO

IDO

2 Mechanisms for PD-L1 upregulation in tumors

TUMOR

Oncogenic

Pathway

(Stat3, Akt)

T CellTCR

PD-1

MHC-pep

PD-L1

Constitutive tumor signaling

induces PD-L1 on tumor cells

Innate (tumor cell intrinsic) Resistance

TUMOR

Stats

TUMORT Cell

T Cell

IFN-g

T cell-induced

PD-L1 up-regulationAdaptive Resistance

2 Mechanisms for PD-L1 upregulation in tumors

TUMOR

Oncogenic

Pathway

(Stat3, Akt)

T CellTCR

PD-1

MHC-pep

PD-L1

Constitutive tumor signaling

induces PD-L1 on tumor cells

Innate (tumor cell intrinsic) Resistance

TUMOR

Stats

TUMORT Cell

T Cell

IFN-gTumor-induced

PD-1 up-regulation TGF-b

Adaptive Resistance

Treg, Mf

TGF-b enhances TCR-dependent transcriptionalinduction of PD-1 in T cells – this effect is mitigated

by a TGF-bRI inhibitor

T cells

CD4cre xSMAD2-flox

aCD3

aCD3+TGF-b1

MFI 343

MFI 951

MFI 248

MFI 338

aCD3

aCD3+TGF-b1

CD4cre xSMAD3-flox

TGF-b dependent induction of PD-1 is dependent on SMAD3 binding to SBE in the PD-1 promoter

Effect of TGF-b signaling in T cells ontumor growth and PD-1 expression on TIL

dnTGF-bRIWTIsotype

Growth of B16 melanoma in wildtype mice vsT cell-specific TGF-bRI KO mice

TGF-bRI KO TIL havemuch lower PD-1

expression

High PD-1 expression on tumor specific TIL is Smad3 dependent

Control Ab Wt TIL Smad3KO TIL

Can a tumor’s genetics

affect it’s microenvironment

and, as follows, its response

to immunotherapy?

Lawrence et al., Nature 2013

Mutational density among cancers sorted by histologyDoes higher mutational load = anti-PD-1 responsiveness?

MSIhi

POLE/POLD1

HPV+

Cancer Genotype linked to Immune Microenviroment – the

case of Colorectal Cancer

• Lymphocyte infiltration into CRC an independent predictor of clinical outcome (Gallon et al)

• Major genetic classification of CRC based on loss of key mismatch repair genes (Microsatellite instable vs stable)

• Even though CRC does not typically respond to anti-PD-1 as monotherapy, one CRC CR was observed – this patient was MSIhi (Lipson et al, CRC 2013)

Analysis of the immune microenvironment of CRC reveals

a subset (~20%) of tumors that are “inflammed”

“No

n-i

nfla

mm

ed”

“In

fla

mm

ed

”

0.1mm

CD8 20x

CD8 20x Foxp3 20x

0.1mm

CD4 20x

CD4 20x

Foxp3 20x

mRNA analysis from LCM reveals that MSI tumor microenvironment has enhanced Th1, CTL and checkpoint expression relative to MSS

MSS

MSITh1 CTL Th17 Treg

Pro-inflammation Checkpoints Metabolism

CD4

PD-L1 expression on myeloid cells inside MSI tumors

MSI

MSS

0.1mm

0.1mm

CD 163 20x PD-L1 20x

CD 163 20x PD-L1 20x

*

*

*

*

MSI

MSS

0.1mm

0.1mm

PD-1 20x LAG-3 20x

PD-1 20x LAG-3 20x

*

*

*

*

PD-1 & LAG-3 expression at TIL and stroma

according to mismatch repair status

MSIhi CRC TIL uniquely contain PD-1hi/IFNghi

populations of CD4 and CD8 T cells

1009080706050403020100

-10-20-30-40-50-60-70-80-90-100

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14

% CEA changefrom initiationof treatment

Weeks from initiation of therapy

CEA changes in response to anti-PD-1 according to mismatch repair status

MSSMSI

Does genetics of colon cancer affect microenvironment? MSIhi vs MSS

No POLE or POLD2 mutation – 49 total nonsynonymous exomic mutations (standard for MSS)

MSS MSS MSS

HLA allelec Gene Peptide lengthd AA Sequence Predicted logIC50e AA sequence Predicted logIC50HLA-A*02:01 IGHV3-48 9 FLVAILEGV 2.47 FLVAILEDV 5.35HLA-A*02:01 IGHV3-48 9 ILEGVQCEV 50.19 ILEDVQCEV 13.82HLA-A*02:01 GORASP1 9 FLDVSGISL 29.15 FLDVSGIFL 21.79HLA-A*02:01 TTC25 9 LIMTGINYL 12.84 LIMMGINYL 22.61HLA-A*02:01 TTC24 9 ALGELPQAL 47.22 ALSELPQAL 23.15HLA-A*02:01 SPAG16 9 TLYGHTDSV 11.05 TLYRHTDSV 29.03HLA-A*02:01 SAMD3 9 ALKDRFKYV 201.26 ALKDHFKYV 46.2HLA-B*15:01 ITGA10 9 LLRGGRRLF 136.45 LLRGGRCLF 50.8HLA-A*02:01 UGT1A6 9 YLKESSFDA 56.25 YLKESCFDA 59.27HLA-A*02:01 BUB1B 9 LLPEEDLDV 301.08 LLPEEYLDV 61.05HLA-B*35:01 RFPL3 9 DAESGSHVY 65.93 DAEGGSHVY 66.56HLA-B*35:01 RRN3 9 VPSTPWFLM 60.45 VPLTPWFLM 156.83HLA-B*35:42 RRN3 9 VPSTPWFLM 60.45 VPLTPWFLM 156.83HLA-B*15:01 ZDHHC23 9 GLFLILLAL 3199.2 GLFLILLAF 234.09HLA-B*15:01 DRD5P2 9 VLKTLSVIM 376.51 VLKPLSVIM 272.65HLA-B*15:01 ZNF729 9 VIHTGEKPC 26452.98 VIHTGEKPY 463.21HLA-B*35:01 PCDHGB6 9 LPDLSDRPV 1077.82 LPDLSDCPV 567.31HLA-B*35:42 PCDHGB6 9 LPDLSDRPV 1077.82 LPDLSDCPV 567.31HLA-B*15:01 UGT1A6 9 KESSFDAVF 425.15 KESCFDAVF 584.76HLA-B*15:01 KIAA0495 9 LPRQPQPQL 18340.52 LLRQPQPQL 672.24HLA-B*35:01 TTC24 9 LAYHALGEL 1558.29 LAYHALSEL 736.02HLA-B*15:01 DLG3 9 KLNGSGPSW 1050.84 KLNGSSPSW 872.01HLA-B*15:01 TTC24 9 LAYHALGEL 2343.15 LAYHALSEL 914.14HLA-B*35:01 PDZD4 9 RSADGKRVY 5016.45 CSADGKRVY 992.94HLA-B*35:01 Myo15b 9 QSLYCRIAL 2123.46 QSLYCHIAL 1192.06HLA-B*15:01 MAP7D3 9 AQKEKFTAI 482.74 AQTEKFTAI 1226.22HLA-B*35:01 SAMD3 9 RALKDRFKY 2059.72 RALKDHFKY 1254.94HLA-B*35:01 ZNF729 9 VIHTGEKPC 34053.93 VIHTGEKPY 1827.87HLA-B*35:01 CLIP1 9 LARDGHDQH 10771.59 LAWDGHDQH 2067.6

HLA-A*02:01 GORASP1 10 SLLDNSNASV 7.51 FLLDNSNASV 2.2HLA-B*15:01 Cpsf4l 10 MQKGTGLLPF 6.3 MQKCTGLLPF 6.08HLA-A*02:01 CMTM4 10 FIASIVLAAL 6.98 FIASIILAAL 6.31HLA-A*02:01 GORASP1 10 FLDVSGISLL 16.47 FLDVSGIFLL 8.52HLA-A*02:01 PCDHGB6 10 ILPDLSDRPV 50.57 ILPDLSDCPV 9.09HLA-A*02:01 IGHV3-48 10 AILEGVQCEV 13.45 AILEDVQCEV 15.64HLA-A*02:01 ZDHHC23 10 FLILLALHRA 37.93 FLILLAFHRA 23.57HLA-B*15:01 IGLV3-16 10 TLCTGSEASY 226.87 TLRTGSEASY 43.87HLA-B*15:01 L1CAM 10 EQSPRRLVVF 413.9 KQSPRRLVVF 73.75HLA-B*35:01 MAGEA11 10 NANGRDPTSY 82.69 NANGRNPTSY 95.51HLA-B*35:01 DRD5P2 10 KTLSVIMGVF 5249.76 KPLSVIMGVF 141.82HLA-B*35:01 SYNPO2L 10 APVPQPLQPG 4068.97 APVPQPLQPW 143.86HLA-B*15:01 ITGA10 10 RLFLSGAPRF 44.4 CLFLSGAPRF 155.24HLA-B*15:01 ITGA10 10 MLLRGGRRLF 238.49 MLLRGGRCLF 160.97HLA-B*15:01 ZNF729 10 KVIHTGEKPC 14292.42 KVIHTGEKPY 201.6HLA-B*35:01 Cpsf4l 10 MQKGTGLLPF 177.85 MQKCTGLLPF 216.73HLA-B*35:01 BUB1B 10 DPERLLPEED 30969.63 DPERLLPEEY 223.11HLA-B*35:01 PCDHGB6 10 LPDLSDRPVL 381.25 LPDLSDCPVL 276.1

HLA-A*02:01 C14orf180 11 YIADKNATATV 6.45 YIADKNTTATV 8.25HLA-B*15:01 UGT1A6 11 YLKESSFDAVF 10.19 YLKESCFDAVF 13.25HLA-A*02:01 CMTM4 11 FLFFIASIVLA 14.8 FLFFIASIILA 16.46HLA-A*02:01 Myo15b 11 SLYCRIALKSL 59.96 SLYCHIALKSL 29.33HLA-A*02:01 PCDHGB6 11 ILPDLSDRPVL 165.79 ILPDLSDCPVL 45.2HLA-A*02:01 DRD5P2 11 VLKTLSVIMGV 131.24 VLKPLSVIMGV 64.22HLA-B*35:01 RFPL3 11 DAESGSHVYTF 71.9 DAEGGSHVYTF 86.74HLA-A*02:01 ITGA10 11 MLLRGGRRLFL 195.59 MLLRGGRCLFL 99.58HLA-B*15:01 Cpsf4l 11 EMQKGTGLLPF 95.1 EMQKCTGLLPF 123.43HLA-B*15:01 IGLV3-16 11 LTLCTGSEASY 108.25 LTLRTGSEASY 128.78HLA-B*35:01 PCDHGB6 11 RPVLSDPQAEL 515.55 CPVLSDPQAEL 135.62HLA-B*35:01 Cpsf4l 11 EMQKGTGLLPF 133.46 EMQKCTGLLPF 156.8HLA-B*15:01 ITGA10 11 SMLLRGGRRLF 279.45 SMLLRGGRCLF 192.45HLA-B*35:01 RHCE 11 YPRSVRRCLPL 242.78 YPRSVRRWLPL 196.7HLA-B*15:01 MAP7D3 11 AQKEKFTAILY 62.3 AQTEKFTAILY 236.62HLA-B*35:01 NBEAL2 11 GPAAEEALNVF 82.97 GPTAEEALNVF 242.2HLA-B*35:01 GORASP1 11 DPGFLDVSGIS 10111.23 DPGFLDVSGIF 244.86HLA-B*15:01 UGT1A6 11 KLVEYLKESSF 79.1 KLVEYLKESCF 299.6

Normal peptidea Tumor/neoantigenb

49 mutations – many potential neoantigens

Normal Tumor0

100

200

300

400

Pre

dic

ted A

*0201 logIC

50

Normal Tumor0

10000

20000

30000

Pre

dic

ted B

*1501 logIC

50

Normal Tumor0

10000

20000

30000

40000

Pre

dic

ted B

*3501 logIC

50

A B CHLA-A201 HLA B1501 HLA-B3501

Potential Neoantigens in MSS patient with 49 coding mutations

De Novo HLA binding

New TCR contact in peptide with moderate – high HLA binding affinity

Therapeutic Implications for PD-1 Pathway

Blockade of Adaptive Resistance Model

Strong Endogenous

Anti-tumor Immune

Response

PD-L1 up-regulation

on tumor/TAMRESPONSE

Single agent

Anti-PD-1

Weak Endogenous

Anti-tumor Immune

Response

No PD-L1 up-regulation

on tumor/TAM

NO

RESPONSE

Single agent

Anti-PD-1

Weak Endogenous

Anti-tumor Immune

Response

Endogenous

Anti-tumor Immune

Response

RESPONSE

PD-L1 expression

on tumor/TAM

Inducer of Anti-tumor Immunity, ie vaccine1

Anti-PD-12

Cyclic Dinucleotides (CDNs) and Intracellular DNA Activate Innate Immunity via STING

• Cyclic dinucleotides (CDNs) regulate

gene expression in bacteria

• Intracellular bacteria secrete CDNs

through MDRs into the cytosol of

infected cells

• cGAS synthesizes c-GMP-AMP in

response to binding cytosolic DNA

• CDNs are sensed by STING (Stimulator

of Interferon Genes)

• STING activates innate immunity by

signaling through the TBK-1/IRF-3 axis

and NF-κB

• The potency of vaccines is significantly

enhanced by co-formulation with CDNs IFN-βPro-Inflammatory

cytokinesNF-KB IRF-3

IKK

Cyclic-di-GMP

c[G(3’,5’)pG(3’,5’)p]

IRF-3

TBK-1

ATP+ GTP

cGAMP

cGAS

Pathogen

CDG

Bacteria Protozoa Viruses

Cyclic-GMP-AMP

c[G(2’,5’)pA(3’,5’)p]

DDX41 IFI16 DAI DNA-PK

Th-2cytokinesSTAT6

STAT6NF-κB

STING

STING-CDN Crystal Structure Enables Rational Design of Molecules with 100x Potency Relative to Native

Molecules – most potent known DC activators

c-di-GMP STING co-crystal

: Phosphodiesterase resistance

CDN composition & modifications

: Phosphate bridge linkage

: Formulation enhancement

: Purine nucleotide base

CDN structure

0.00

0.20

0.40

0.60

0.80

1.00

7 12 17 22 27

Tum

or

Vo

lum

e

(cm

3)

GM-vaccineSTINGVAXSTINGVAX+Anti-PD1

Untreated

GM-Vaccine

STINGVAX (di-cAMP)

STINGVAX (R,R-di-A)

STINGVAX (ML-di-A)

CD8+INFg+ PD-L1

STINGVAX vaccination results in enhanced IFNgproduction in tumors associated with increased PD-L1

induction and synergy with anti-PD-1

Combination checkpoint blockade

LAG-3 KO

WT

PD-1 KO

PD-1/LAG-3

DKO

LAG-3 / PD-1 Double KO Mice Reject

Poorly Immunogenic B16 Tumors

Relative Intensity

Log

2Fold C

hange

Up

-re

gula

ted

in P

d-1

Hi v

sLo

Up

-re

gula

ted

in P

D-1

Lo

vs

Hi

Expression profiling

of PD-1hi vs PD-1lo

TIL reveals additional

checkpoint pathways

targetable with Ab:

LAG-3

Tu

mo

r V

olu

me

m

m3

Control Ab

(0/10 TF)Anti-

LAG-3

(0/10 TF)

Anti-PD-1

(4/10 TF)

0 10 20 30 40 50

0

500

1000

1500

2000

Anti-

LAG-3 +

Anti-PD-

1

(8/10 TF)

Days Post-Treatment Initiation

Tum

or

Volu

me

Epigenetic modulation of tumor antigenicity and the tumor

microenvironment

Multiple immune genes induced by 5-aza treatment

Checkpoint receptors

Interferon receptors/Stats

Antigen Processing/Presentation

Interferon pathway

IFN Regulatory Factors

Cancer Testes Ag

Checkpoint ligand

Antigenprocessing& presentation

STATs/IFNgR

Interferonresponsegenes

BATF familytranscriptionfactors

InterferonRegulatoryfactors

Epigenetic silencing of genes in innate and adaptive immunity in NSSCL

AdenocarcinomaSquamous cell ca

Response to anti-PD-1/PD-L1 after5’aza-CR + HDACi in 5 pt with NSCLC

Stable disease

Objective Response

% S

urv

iva

l

TimeControl

Standard or Other Therapy

Checkpoint Blockade

Combination

Improving Survival with Combination Therapy

Sponsored by Bloomberg Foundation, Barney Foundation,

Merieux Foundation, SU2C and Melanoma Research Alliance

Suzanne Topalian Janis TaubeJulie Brahmer Franck HousseauCharles Drake Nico LlosaBill Sharfman Ben ParkBob Anders Cindy SearsBert Vogelstein Young KimKen KinzlerNick PapadopolusSteve Baylin