unleashing the immune system against cancer

Targeted immunotherapy: unleashing
the immune system against cancer
Suzanne L. Topalian, MD
Johns Hopkins University School of Medicine
Sidney Kimmel Comprehensive Cancer Center
2013 MMS Annual Education Program:
Navigating the Currents of Change
Disclosure Information
Suzanne L. Topalian, MD
I have the following financial relationships to disclose:
– Consultant for: Bristol-Myers Squibb (uncompensated), Jounce
Therapeutics, Sanofi, and Amplimmune (spouse)
– Research grant support from Bristol-Myers Squibb
– Royalties through institution: Amplimmune (spouse), Bristol-Myers
Squibb (spouse)
“Cancer is a genetic disease”
rationale for personalized medicine
limitation: rapid resistance
Cancer is an immunological disorder
immunotherapy is a common
denominator that can activate immune
responses against mutant proteins
the adaptable immune system can keep
pace with tumor evolution
The adaptive immune system:
an “ideal” anti-cancer agent
Diversity
T cells - 1018
Antibodies - 1022
Specificity
Can distinguish minute
chemical alterations
Tumor cell
T cell
Memory
After effective antigen
priming, immunity can
last for decades
Immune tolerance: Multiple barriers to
tumor elimination
Regulatory immune cells
T reg cells
MDSCs
T-cell inhibitory receptors
CTLA-4, PD-1, LAG-3, etc.
Immunosuppressive cytokines
IL-6, IL-10, TGF-ß, VEGF
Solution: adoptive
transfer of T cells
optimized/engineered
ex vivo
Solution: systemic blockade
of suppressive cytokines or
inhibitory receptors/ligands
(“immune checkpoints”)
Advances in Immunotherapy 2011
Anti-CTLA-4
(ipilimumab) was
approved for treating
metastatic melanoma
based on improved
overall survival in a
randomized study.
However, the grade 3-4
drug-related toxicity rate
approximated the clinical
benefit rate
Adapted from Hodi et al, NEJM 2010
Ipilimumab
Control
Antigen presenting or tumor cell
PD-L1 is expressed
by many human
tumors and may play
a pivotal role in
local cancer
immunosuppression
T cell
The PD-1/PD-L1
pathway:
New strategies
for immune
checkpoint
blockade
Pardoll 2012
Role of PD-1 in suppressing antitumor
immunity
Activation
(cytokines, lysis, prolif., migration)
APC
T cell
B7.1
MHC-Ag
CD28
TCR Signal 1
(-)
(-)
(-)
PD-1
PD-L1
Inhibition
Tumor
(anergy, exhaustion, death)
Tumor
Role of PD-1 in suppressing antitumor
immunity
Activation
(cytokines, lysis, prolif., migration)
APC
T cell
B7.1
MHC-Ag
CD28
TCR Signal 1
(-)
(-)
(-)
AntiPD-1
PD-1
PD-L1
Inhibition
Tumor
(anergy, exhaustion, death)
Tumor
CTLA-4 vs. PD-1: Distinct immune checkpoints
B7.1/2
APC
CD28
signal 1
CTLA-4 to
surface
APC
Costim.
ligand Costim. receptor
signal 1
Traffic to
periphery
signal 1
CTLA-4
Naïve/resting T cell
APC
+
Experienced
T cell
Tissue
PD-L1
T cell priming
Topalian et al., Curr Opin Immunol 2012
signal 1
PD-1
Nivolumab (anti-PD-1)
(BMS-936558, MDX-1106, ONO-4538)
 Fully human IgG4 anti-human PD-1 blocking mAb
 High affinity for PD-1 (KD ~ 3 nM), blocks binding to
both known ligands: PD-L1 (B7-H1, broadly
inducible) and PD-L2 (B7-DC, selective for APCs)
 Favorable safety profile and preliminary evidence of
clinical activity in patients with treatment-refractory
solid tumors in a first-in-human study (Brahmer et al.,
J Clin Oncol 2010)
Two early phase clinical trials of anti-PD-1 (nivolumab)
in patients with advanced solid tumors
(Brahmer et al., JCO 2010; Topalian et al., NEJM 2012)
1st Treatment Cycle
Day 1
Dose 0.3-10 mg/kg
Day 57
Scans
43
Day 85
Scans
2 years
or until CR/PR or PD
Follow Up or Additional
Treatment Cycle(s)
1st Treatment Cycle
Days 1, 15,
29
Dose 0.1-10 mg/kg
Follow Up or Additional
Treatment Cycle(s)
Day 57
Scans
2 years
or until CR or PD
Eligible patients: treatment-refractory metastatic lung cancer,
melanoma, kidney, colon, or prostate cancer
CRC
Durable anti-PD-1 responses OFF THERAPY:
immune memory
CR Stop Rx
Latest eval.: CR
(Lipson et al., CCR 2013)
0
1 yr
RCC
Stop
Rx Best resp.(PR)
0
1 yr
1 yr
3 yr
CR
2 yr
Stop Best
Rx resp.(PR)
MEL
0
2 yr
2 yr
4 yr
Latest eval.: CR
3 yr
4 yr
5 yr
New LN
met
Resume
Rx
PR
Latest eval.
PR
3 yr
4 yr
5 yr
Clinical activity of anti-PD-1 (nivolumab)
ORR (CR/PR) SD 24 wk
No. pts (%) No. pts (%)
Tumor
Type
Dose
(mg/kg)
No.
pts
NSCLC
1-10
122
20 (16)
11 (9)
MEL
0.1-10
106
33 (31)
6 (6)
RCC
1 or 10
34
10 (29)
9 (27)
 294 patients started therapy between 2008-2012 and had ≥6
mo. follow-up. No ORs in CRC (n=19) or CRPC (n=17)
 28/54 responses lasted ≥1 yr in patients with ≥1 yr follow-up
 Toxicity: grade 3-4 drug-related AEs15%, mortality 1%.
Topalian et al., ESMO 2012
Change in Target Lesions from Baseline (%)
Anti-PD-1 response kinetics and durable tumor
control off therapy
100
Melanoma 1 mg/kg
90
80
70
60
50
40
30
20
10
First occurrence
of new lesion
Maximum
duration of
therapy
On study
Off study
0
-10
-20
-30
-40
-50
-60
-70
-80
-90
-100
0
10
20
30
40
50
60
70
80
90
100
Weeks Since Treatment Initiation
110
120
130
140
Partial response of locally advanced primary
melanoma to anti-PD-1
Pre
Tumor
Lymphs
2 mo.
 35-year-old patient had disease progression after surgery and IL-2.
 Response to anti-PD-1 ongoing at 23 months.
Partial regression of metastatic kidney cancer
in response to anti-PD-1
Pretreatment
6 months
 57-year-old patient developed progressive disease after receiving
sunitinib, temsirolimus, sorafenib, and pazopanib
 Currently in cycle 12 anti-PD-1 therapy (~23 months) with ongoing PR
Response of a “non-immunogenic” tumor to
anti-PD-1: Squamous cell lung cancer
Pre-Rx
2 months
12 months
History:
61-year-old patient
with stage IV
NSCLC
refractory to
multiple surgeries,
RT, 2 multidrug
chemotherapy
regimens, and
epigenetic therapy
(5-AZA and
entinostat).
Recently completed
2 yrs anti-PD-1
therapy.
Drug-related AEs of special interest in ≥ 3% of
304 patients receiving anti-PD-1
AEOSI (“irAE”)
All grades (n, %)
Grade 3-4 (n, %)
Any
138 (45)
18 (6)
Rash
41 (14)
0
Diarrhea
36 (12)
3 (1)
Pruritis
31 (10)
1 (0.3)
ALT 
13 (4)
2 (1)
AST 
11 (4)
2 (1)
TSH 
11 (4)
1 (0.3)
Pneumonitis*
10 (3)
3 (1)
Vitiligo
9 (3)
0
Infusion reaction
8 (3)
0
*There were 3 (1%) deaths in patients with pneumonitis (2 NSCLC, 1 CRC).
AEOSIs in ≤1% of pts included colitis, hepatitis, hypophysitis, nephritis and
thyroiditis. Analysis as of July 2012.
Topalian et al, ESMO 2012
Partial response of metastatic mucosal melanoma
to anti-PD-1, associated with nephritis
Pre-Rx
8 months
Prior therapies
included multiple
surgeries, RT and
temozolomide.
Nephritis
developed after 8
months of antiPD-1 therapy,
associated with
administration of
radiographic
contrast dye.
Nephritis associated with anti-PD-1 therapy:
serologic and cellular immune components
H&E, 400 X
CD3
tubule
glomerulus
plasma
cell
intraepithelial
T cell
B7.1
PD-1
? antiapoptotic
signal in
tumor cells
PD-L1
PD-L2 Anti-PD-1
Anti-PD-L1
? antiapoptotic
signal in
tumor cells
(-) signal
B7.1
PD-1
PD-L1
PD-L2
(-) signal
If PD-1 pathway blockade acts locally in the
tumor immune microenvironment……….
Then it follows that the tumor site may hold
the key to optimizing anti-PD-1/PD-L1 therapy:
Biomarkers to select patients most likely
to respond to therapy
Early indicators of treatment outcomes
Identification of resistance pathways and
development of synergistic combinatorial
therapies
Melanoma
Preliminary molecular marker
studies: Correlation of PD-L1
expression in pretreatment
tumor biopsies with clinical
response to anti-PD-1
Proportion of patients
18/18
1
*
0.8
0.6
18/31
Kidney
PD-L1(+)
CR/PR
PD-L1(-)
NR
13/31
0.4
p=0.002
Lung
0.2
0/18
0
PD-L1(+)
PD-L1(-)
49 patients include 20 with melanoma,13
NSCLC, 7 colon, 6 kidney, and 3 prostate
cancer.
* Normal renal glomerulus
Correlation of PD-L1 expression with tumor type
in 49 patients treated with anti-PD-1
18
No. patients
16
14
12
10
PD-L1(+)
8
PD-L1(-)
6
4
2
0
0
MEL
Responders/total: 8/20
NSCLC
3/13
RCC
2/6
CRC
0/7
CRPC
0/3
Patients were “PD-L1+” if ≥5% of tumor cells in any tumor biopsy expressed
cell surface PD-L1, using mAb 5H1 (L. Chen) and manual staining technique.
2 Mechanisms for PD-L1 up-regulation in tumors
Innate Resistance
MHC-pep
TCR
TUMOR
T Cell
Constitutive tumor signaling
induces PD-L1 on tumor cells
Oncogenic
Pathway
PD-L1
PD-1
Adaptive Resistance
TUMOR
T Cell
PD-L1 expression
reflects immune reaction
TUMOR
T Cell
Stats
IFN-g
Focal PD-L1 expression in melanoma:
geographic co-localization with TILs creates a “shield”
against immune attack
Tumor
Lymphs
PD-L1 expression in melanoma correlates with the
presence of TILs and increased overall survival in
patients with metastatic disease
% of cases
80
60
TILs
PD-L1(+)
PD-L1(-)
No TILs
p<0.0001
40
20
% survival
100
p=0.032
0
PD-L1(+), n=57
PD-L1(-), n=93
Duration follow-up (mo)
TILs are necessary but not sufficient for PD-L1 expression
(Taube et al., Science Transl Med 2012)
PD-L1
PD-L1 expression on tumor cells and TAMs
in oropharyngeal SCCHN: association with
CD8+ TILs, IFN-g and HPV+ tumors
tumor
p=0.002
CD68
# patients
CD3
p=0.003
15
10
5
0
PD-L1+
PD-L1-
HPV+
(n=20)
HPV(n=7)
Lyford-Pike, Pai , et al., Cancer Res 2013
PD-L1
tumor
PD-L1 expression on tumor cells and TAMs
in Merkel cell Ca: association with
CD8+ TILs, MCV, and overall survival
CD68
CD8
PD-L1+
10
0
0
MCV+
(n=34)
PD-L1p=0.01
MCV(n=8)
Survival probability
stroma
# Tumors
20
Duration of follow-up (years)
Lipson, Taube, et al., Cancer Immunol Res 2013
PD-1 pathway in virus-associated cancers:
at the crossroads of cancer immunology
and microbial immunology
 Principles of PD-1/PD-L1 mechanism-of-action first
demonstrated in preclinical models of infectious disease
 PD-1 pathway restrains host immune responses against
chronic infection
 Virus-associated cancers account for a large proportion
of cancer deaths worldwide
 PD-1 blockade now in the clinic for virus-associated
hepatocellular cancer, in planning for HIV
Potential PD-1 pathway interactions
in chronic HIV infection
 Increased PD-L1
expression on APCs
in HIV infection
PD-L1/2
PD-1
??
CD4 T cell
HIV+
APC
CD4 T cell
HIV+
??
 PD-1 and other coinhibitory molecules
are expressed on HIV
specific CD8 T cells
(-)
(-)
(-)
 PD-1+ CD4 T
cells might be an
important latent
viral reservoir
CD8
T cell
Dissecting mechanisms and resistance pathways:
Immune infiltrates at the boundary of
PD-L1+ melanoma cells
PD-L1
H&E
H&E
IFN-g mRNA is overexpressed in PD-L1(+) vs. PD-L1(-)
melanomas (Taube et al., Science Transl Med 2012)
 Gene expression profiling reveals inflammatory signature in
PD-L1(+) tumors
Gene expression profile of TILs from
PD-L1(+) vs. (-) melanomas reveals functional
groups of differentially expressed genes and
potential bypass pathways
-log10 p value
PRF1
3.1x
IL-10
3.3x
CCL5
2.6x
PD-1
7.3x
TLR3
4.6x
CXCL1
8.3x
CD8a
2.4x
LAG-3
10.7x
Th1
IFN-g
11.3x
IL-22R
-4.4x
IL-21
4.1x
PD-L1
4.4x
log2 fold change
multiplex qRT-PCR, CD45 normalization
CD8 T
Checkpoint
Association of lymphocyte activation gene-3
(LAG-3)+ lymphocytes with
PD-L1+ melanoma cells
H&E
PD-L1
LAG-3
IHC
tumor
lymphs
ISH
Young, Taube et al., AACR 2013 abstr. #465
Meeker, Taube, Advanced Cell Diagnostics
Synergistic anti-tumor effects of dual
checkpoint blockade: anti-PD-1 and
anti-LAG-3 in a murine tumor model
AntiLAG-3
(0/10 TF)
2000
1500
1000
Anti-LAG-3
+Anti-PD-1
(8/10 TF)
500
50
40
30
20
0
10
Anti-PD-1
(4/10 TF)
0
Tumor Volume (mm3)
Control Ab
(0/10 TF)
Days Post-Treatment Initiation
(Shown: MC38 colon cancer)
Drake, Vignali, Korman, Pardoll et al., Cancer Res 2012
Therapeutic implications for PD-1 pathway
blockade in adaptive resistance model
Anti-PD-1 monotherapy
Strong endogenous
anti-tumor immune
response
PD-L1 up-regulation
in tumor
Weak endogenous
anti-tumor immune
response
No PD-L1 up-regulation
in tumor
RESPONSE
1 Inducer of anti-tumor immunity (vaccine, TKIs, “immunogenic” chemo, RT)
Endogenous
anti-tumor immune
response
PD-L1 expression
in tumor
2
Anti-PD-1
RESPONSE
PD-1 pathway blocking agents in clinical testing
www.clinicaltrials.gov
Target Molecule
PD-L1 (B7-H1)
Company
PD-1
Amplimmune/GSK
AMP-224 (PD-L2/Fc fusion
protein)
N/A
Bristol-Myers Squibb Nivolumab/BMS-936558/
MDX-1106/ Ono-4538
(fully human IgG4 mAb)
BMS-936559/MDX1105
(fully human IgG4 mAb)
CureTech
Pidilizumab/CT-011
(humanized IgG1 mAb)
N/A
Genentech
N/A
MPDL3280A
(Fc-modified IgG1 mAb)
MedImmune
N/A
Medi4736 (mAb)
Merck
Lambrolizumab/MK-3475
(humanized IgG4 mAb)
N/A
CTLA-4
Demonstration
that B7.1/2 are
CTLA-4 ligands
Demonstration
of potent immune
regulatory role of
CTLA-4 in KO mice
Cloning of
CTLA-4 gene
1985
Treatment of
murine tumors
with anti-CTLA-4
1990
Cloning of
PD-1 gene
PD-1
1995
First demonstration
of clinical activity of
anti-CTLA-4 in melanoma
Anti-CTLA-4
confers survival
benefit in
melanoma
Anti-CTLA-4
into patients
2000
Demonstration
of organ-specific
autoimmunity
in PD-1 KO mice
2005
PD-L1 expression
in murine tumors
confers immune
resistance & high
PD-L1 expression
in human tumors
PD-L1/PD-L2
Identified as
PD-1 ligands
FDA approval
for anti-CTLA-4
in melanoma
2010
2015
First demonstration
of clinical activity of
anti-PD-1
Anti-PD-1
into patients
Predicted
FDA approval
of anti-PD-1
Anti-PD-1 & anti-PD-L1
Induce durable tumor
responses in melanoma,
kidney & lung cancer
Pardoll , Nature Immunol 2012
How can the path to clinical approval
be accelerated?
Regulatory challenges for targeted immunotherapies:
 Defining appropriate endpoints for clinical
efficacy
 Understanding unique side-effects and
developing effective monitoring and
management guidelines
 Identifying biomarkers predicting response,
which may be dynamic and tissue-specific
ACKNOWLEDGEMENTS
Thanks
to collaborating
clinical
trial centers.
Sponsored by
BMS, NIH,
Barney Foundation,
and Melanoma
Research Alliance
Sponsored by BMS, NIH, Barney Fdn., and Melanoma Research Alliance