cit: from translational research to clinical application...van rooij, van buuren jco 2013 cancer...

CIT: from translational research to clinical application

John Haanen

ESMO I-O Preceptorship Zürich 2017

Disclosures

• I have provided consultation, attended advisory boards, and/or

provided lectures for: Pfizer, MSD, BMS, IPSEN,

Roche/Genentech, NEON Therapeutics, Novartis for which NKI

received honoraria

• Through my work NKI received grant support from BMS, MSD,

Novartis

• I declare no conflict of interest

Cancer Immunotherapy

…fighting cancer but ignoring the tumor…

unleashing the immune system or

harnessing the immune system

to combat cancer

2013 2014 2015

What is the science behind CIT?

van Elsas et al., J Exp Med 1999

van Elsas et al., J Exp Med 1999

Prevention of outgrowth and rejection of aggressive B16 melanoma using CLTA-4 blockade and vaccine

Combination of anti-CTLA-4 + vaccine leads to rejectionof pulmonary metastases and immune infiltrates

van Elsas et al., J Exp Med 1999

Melanoma bearing mice successfully treated withanti-CTLA4 and vaccine develop vitiligo

van Elsas et al., J Exp Med 1999

Cancer immunity cycle

Chen & Mellman Immunity 2013

CTLA4

PD1/PDL1CTLA4

T cell signaling

T cell activating and inhibiting molecules and signals

Melero, Haanen. Nat Rev Cancer 2015

Checkpoint molecule CTLA4

CTLA4 plays a role during T cell priming

Ribas. N Engl J Med 2012

Human CTLA4 blockade

• Ipilimumab, IgG1 human mAb, selected for its lack of ADCC/CDC activity– Administered every 3 weeks x 4, 3 mg/kg– Showed improvement in mOS of 4 months in

metastatic melanoma– Approved in 2011

• Tremelimumab, IgG2 human mAb– Administered every 3 months, 15 mg/kg– Failed to show improvement in mOS

Efficacy of ipilimumab as first line treatment

Robert et al., NEJM 2015

ORR: 11.9%

CTLA-4 blockade (ipilimumab) can induce long-term survival (pooled overall survival analysis including Expanded Access Program data from 4846 patients)

4846 1786 612 392 200 170 120 26 15 5 0

Median OS (95% CI): 9.5 months (9.0–10.0)

3-year OS rate (95% CI): 21% (20%–22%)

Pro

port

ion

Aliv

e

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Months0 12 24 36 48 60 72 84 96 108 120

Ipilimumab

CENSORED

Patients at RiskIpilimumab

Schadendorf D, et al. J Clin Oncol 2015

Treatment with anti-CTLA-4 mAb

Maker et al., Ann Surg Oncol 2005

Auto-immune uveitis afteranti-CTLA-4 treatment

Aftertreatment

Immune related adverse events upon anti-CTLA-4 mAb treatment

colitis hypophysitis

How does CTLA4 blockade lead to anti-tumor immune responses?

Two not mutually exclusive mechanisms have been proposed:1. Priming of tumor-reactive T cells

– Against shared tumor associated antigens– Against mutated (neo) antigens

2. Depletion of regulatory FoxP3+ T cells from the tumor microenvironment

1. Self antigens (to which tolerance is incomplete)Shared between patients

2. ‘Neo-antigens’, epitopes that arise as a consequence of tumor-specific mutationsIn large part patient-specific, hence generallyignored

What could tumor-specific cytotoxic T cells detect on human cancer?

Kvistborg et al., Science Transl Med 2014

Two models proposed

Analysis of PBMC from 40 ipilimumab treated melanoma patients for 75 tumor associated

antigens

Kvistborg et al., Science Transl Med 2014

Generation of pMHC multimers by UV-induced peptide ex change

Peptide 1 Peptide 2 Peptide 3

Disintegration

Rescue Peptides

Toebes et al. Nat. Med. 2006Bakker et al. PNAS 2008

Allows generation of 1000s of pMHC in parallel

Generate fluorochromeconjugated MHC multimers

Mix to create a collectionof differentially encodedMHC multimers

Assembly of combinatorialcodes on T cell surfaces

Analysis by flow cytometry

T cell T cell T cell

Self-assembling molecular codes

Allows detection of 47 T cell responses in parallel

Hadrup et al, Nat Methods 2009, Andersen et al, Nat Prot 2012.

Flow results…

Kvistborg et al., Science Transl Med 2014

Ipilimumab treatment leads to broadening of the anti cancer IR

Kvistborg et al., Science Transl Med 2014

Kvistborg et al., Science Transl Med 2014

T cell responses against shared tumor antigens

1. Self antigens (to which tolerance is incomplete)Shared between patients

2. ‘Neo-antigens’, epitopes that arise as a consequence of tumor-specific mutationsIn large part patient-specific, hence generallyignored

What could tumor-specific cytotoxic T cells detect on human cancer?

Generate map of tumor-specific mutations (ExomeSeq)

Determine which mutatedgenes are expressed (RNASeq)

Predict epitopes for each mutation/ each HLA-allele in silico

Screen for T cell recognitionof mutated epitopes

MDLVLNELVISLIVESKLLEHLA-A2HLA-B7HLA-C2

T cell

Analyzing the neo-antigen-specific T cell repertoire in human cancer?

Pt 002: Partial response upon anti-CTLA4 treatment

pre-treatment post-treatment

Van Rooij et al., J Clin Oncol 2013

Analyzing the neo-antigen-specific T cell repertoire in human cancer?

T > A

/ A >

TT >

G /

A > C

T > C

/ A >

GC >

A /

G > T

C > G

/ G >

CC >

T /

G > A

0

20

40

60

80

100

Mutation Type%

of

No

n-S

yno

nym

ou

s m

uta

tio

ns

(n=1036)Resected tumor material

Isolate tumor cells

Identify tumor-specific mutations

Analyzing the neo-antigen-specific T cell repertoire in human cancer?

Screen with MHC multimer technology

Resected tumor material

Isolate tumor cells Isolate tumor-infiltrating T cells

Identify tumor-specific mutations

Predict potential epitopes

pMHC PE-Cy7 (ATRS>L)

pMH

CQ

D 6

25 (

AT

RS

>L)

3.3%

Strong T cell response against an ATRS>L neo-epitope within the tumor

Screen with MHC multimer technology

Resected tumor material

Isolate tumor cells Isolate tumor-infiltrating T cells

Identify tumor-specific mutations

Predict potential epitopes

Increased magnitude of neo-antigen-specific T cell response under anti-CTLA4

pMHC PE-Cy7 (ATRS>L)

pMH

CQ

D 6

25 (

AT

RS

>L)

3.3%

van Rooij, van Buuren JCO 2013

Cancer exome-guided immunomonitoring

� Exome-based analysis of neo-antigen specific T cell responses in human

cancer is feasible

- Dissect the role of neo-antigen specific T cell reactivity in melanoma

immunotherapy (TIL therapy, anti-CTLA4, anti-PD1 etc.)

� The T cell based immune system commonly interacts with the consequences

of DNA damage in human melanoma

in 10 out of 12 pts CD8 neo-ag spec TIL

in 5 out of 6 CD4 neo-ag spec TIL

Snyder et al. NEJM 2014

Mutational load and outcome to ipi

Freeman & Sharpe. Nat Immunol 2013

PD1 and PD-L1 checkpoint

Programmed Death-1 receptor (PD1)

• Discovered in 1992 by Honjo and coworkers– Upregulated gene in relation to apoptosis

• Member of the Ig superfamily• Cytoplasmic domains with ITIM and ITSM

– Recruites phosphatases– Inhibits PI3K and AKT activity

• Inducibly expressed by CD4 and CD8 T cells, NKT cells, B cells, monocytes and subtypes of DC

• Expressed by both effector and regulatory T cells• PD1/PD-L1 interaction involved in tolerance and chronic

inflammation• PD1/PD-L1 contributes to functional T cell exhaustion during

chronic infection and cancer

PD-1 pathway inhibits T cell response directlydownstream of the TCR

Freeman PNAS 2008

PD1/PD-L1 play a role at the tumor/effector phase

Ribas. N Engl J Med 2012

Checkpoint molecules PD-1/PD-L1

• Nivolumab: anti-PD1• Pembrolizumab: anti-PD1

• Atezolizumab: anti-PDL1• Durvalumab: anti-PDL1• Avelumab: anti-PDL1

Expression of PD1 ligands

PD-L1 (B7-H1) PD-L2 (B7-DC)

Hematopoietic cells DC, macrophages, B cells, T cells, BM-derivedmast cells

DC, macrophages, B cells, Th2 cells, BM-derived mast cells

Non-hematopoietic cells Vascular endothelium, epithelia, muscle, liver, pancreatic islets, placenta, eye

Few, airway epithelia

Stimuli Interferons (α, β, γ) IL-4 + GM-CSG

Binding partners PD1, B7.1 PD1

Expression by tumors Melanoma, RCC, HNSCC, ovary, NSCLC

PD-L1 on human tumor cells mediatesT cell inhibition

Pardoll DM, Nat Rev Cancer 2012

Expression of PD-L1 co -localizes withTILs

Taube et al. Science Transl Med 2012

PembrolizumabBaseline: April 13, 2012

Images courtesy of A. Ribas, UCLA.72-year-old male with symptomatic progression after bio-chemotherapy, HD IL-2, and ipilimumab

April 9, 2013

Ribas A ASCO 2013

Design of CheckMate 066

Results of the CheckMate 066

Nivolumab improves PFS and OS compared to dacarbazine

Are there any predictive markers for PD1/PDL1 blockade?

Evolution of CD8+ T-cells, according to treatment outcome

IHC Analysis of CD8+ T-cells in samples obtained before and during anti-PD1 treatment

Tumeh et al. Nature 2014

PD-L1 expression and response in melanoma

Pembrolizumab

ORR – “PDL1+”1% cut-off: 49%

10% cut-off: 52%

ORR – PDL1-1% cut-off: 13%

10% cut-off: 23%

Melanoma

Daud, AACR 2014Ghandi, AACR 2014

Mutational burden and clinical benefit from anti-PD1 in NSCLC

Rizvi et al., Science 2015

A neo-antigen specific T cell response induced during anti-PD1 treatment

Rizvi et al., Science 2015

Anti-PD1 in mismatch repair deficient tumors

Le et al., NEJM 2015

A neo-antigen repertoire may only be frequent in so me human cancers

Formation of neo-antigens

Generally

Regularly

Occasionally

Caveat: It is possible that a ‘neglected’ repertoire of neo-antigens existsWe may be underestimating…

Schumacher and Schreiber Science 2015

Checkpoint blockade

Paradigm shift: Cancer immunotherapy can be applied broadly

Requirement for effective CIT

• High mutation burden increases chance for neo-antigen specific T cell response

• T cell infiltrate in tumor or invasive margin– CD4 and/or CD8

• PD-L1 expression (indicative of T cell-tumor interrogation) favors response

• IFN-γ gene signature favors response to CIT

How do we see the future of CIT?Combine strategies that enhance T cell activity with strategies

that direct this T cell activity towards cancer neo-antigens

Schumacher and Schreiber Science 2015

Tumor destruction+

Checkpoint blockade

Neo-ag vaccine+

Checkpoint blockade

Neo-ag specific T cell product+

Checkpoint blockade

Target cancer immunity cycle obstructions

Marit van Buuren

Sander Kelderman

Carsten Linneman

Laura Bies

Daisy Philips

Mireille Toebes

Pia Kvistborg

Maarten Slagter

Joost Beltman

Clinical translation

Nienke van Rooij

Bianca Heemskerk

Raquel Gomez

Joost van den Berg

Maaike van Zon

Noor Bakker

Christian Blank

Chemical Biology

Huib Ovaa

PDX models

Kristel Kemper

Daniel Peeper

Sanger Institute

Sam Behjati

Mike Stratton

AIMM Therapeutics

Remko Schotte

Hergen Spits

LUMC

Els Verdegaal

Sjoerd van der Burg

Cancer Immunotherapy Dream Team

MSKCC

Naiyer Rizvi

Matt Hellman

Alexandra Snyder

Vlad Makarov

Jonathan Havel

Taha Merghoub

Jedd Wolchok

Tim Chan

Utrecht University

Can Keşmir

MD Anderson

Laszlo Radvanyi

Chantale Bernatchez

Patrick Hwu