Upload
others
View
6
Download
0
Embed Size (px)
Citation preview
Melanoma and introduction to immunotherapy
Prof. Olivier Michielin, MS, MD-PhD
Head of Personalized Analytical Oncology,
Department of oncology, CHUV, Lausanne
Swiss Institute of Bioinformatics, Lausanne
March 25th 2019, Berlin, GermanyESMO/ESO Masterclass in Clinical Oncology
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Targeted therapies
Two simultaneous revolutions in stage IV melanoma
Immune therapies
1 Chapman & al. NEJM 2011; 2 Hauschild & al. Lancet 2012; 3 Hodi & al. NEJM, 2010; 4 Flaherty, ASCO 2016; 5 Hodi, AACR 2016; 6 Postow, AACR 2016
2nd Generation I-O: • PD-1 blockade5
• Combined CTLA-4 + PD-16
2nd Generation TKI: • Dual BRAF + MEK
inihibition4
BRAFi1,2
Early clinical benefit
CTLA-4 blockade3
Late clinical benefit
Early and lateclinical benefit
• 2nd generation strategies can both deliver early and late clinical benefits, challenging treatment plans
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Overview of stage IV outcome
1 Ugurel, EJC 2017
• Targeted therapies provide better early outcome…
• … but immuno-oncology curves are crossing at around 14 months 1 …
• … and the difference seems to increase with time
March 25th 2019, Berlin, GermanyESMO/ESO Masterclass in Clinical Oncology
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
What is the best endpoint to describe such a benefit?
1 Asierto, Lancet Oncology 2017
• mOS or mPFS are not adequate to describe the long term benefit
• P. Ascierto and G. Long have proposed to use landmark PFS 1
• PFS is not dependent on subsequent lines
• Another endpoint to drive stage IV development is CR rate…
March 25th 2019, Berlin, GermanyESMO/ESO Masterclass in Clinical Oncology
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Overview of therapeutic options in melanoma
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Overview of therapeutic options in melanoma
Surgery Radio-therapy
Chemo-therapy
Immuno-therapy
Targeted-therapy
Stage I Standard No! No! ? ???
Stage II Standard No! No! Ongoing! ??
Stage III Standard Option?Cave!
No! Standard Standard
Stage IV Option forselected
Option forpalliation
Option Standard Standard
Ad
juva
nt
Me
tasta
tic
Loco-regional Systemic
A multidisciplinary team is needed specially for early stage disease!
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Brief introduction to cancer immunotherapy
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Escaping the immune system is a hallmark of cancer!
Hanahan & Weinberg, Cell 2011
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
CD8+
TCR
MHC-I
Tum
or
cell
Lym
ph
ocy
te
Cytotoxic T Lymphocyte: Shortman, Brunner, Cerottini, J.Exp. Med. 1972
Demonstration of the activity of cytotoxic T lymphocytes
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Various types of tumor antigens
CD8+
TCR
MHC-I1) Differentiation2) Overexpression3) Cancer-Testis4) Mutations
(Neo-antigens)
P53Mutation
P53 Antigenin MHC-1
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Nivolumab, pembrolizumab,
and atezolizumab approval in NSCLC
Nivolumab approval in RCC
Nivolumab and Pembrolizumab
approval in SCCHN
Nivolumab and Atezolizumab
approval in Bladder cancer
Nivolumab and pembrolizumab
approval in Melanoma
PD-1/PD-L1 inhibition successes across multiple tumors
1000
100
10
1
0.1
0.01
n=22
Rh
abd
oid
tu
mo
r
Ewin
g sa
rco
ma
Thyr
oid
AM
L
Me
du
llob
last
om
a
Car
cin
oid
Neu
rob
last
om
a
Pro
stat
e
CLL
Low
-gra
de
glio
ma
Bre
ast
Mu
ltip
le m
yelo
ma
Pan
crea
s
Kid
ney
cle
ar c
ell
Kid
ney
pap
illar
y ce
llO
vari
an
Glio
bla
sto
ma
Cer
vica
l
DLB
CL
Hea
d a
nd
nec
k
Co
lore
ctal
Eso
ph
agea
lad
eno
carc
ino
ma
Sto
mac
h
Bla
dd
er
Lun
g A
D
Lun
g SQ
Mel
ano
ma
20 52 134 26 23 81 227 91 57 121 13 63 214 11 394 219 20 49 181 231 76 88 35 335 179 121So
mat
ic M
uta
tio
n
Freq
uen
cy (
/Mb
)1
1. Lawrence, Nature. 20135t
h ESO-E
SMO L
atin
Amer
ican
Mas
terc
lass i
n Clin
ical O
ncolo
gy
Inflamed
Immune excluded
Immune desert
Immune phenotypes andimmunotherapy
Adapted from• Chen & Mellman, Immunity, 2013• Kim & al. Ann Oncol, 2016
1
Antigenrelease
2Antigenpresentation
3Priming &activation
4
Trafficking
5 Tumor infiltration
6 Tumor recognition
7
Tumor killing
Tumor cells
No T cells
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Immunotherapy:Checkpoints inhibitors
αCTLA-4
αPD-1
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Checkpoints act at different stages of the immune cycle
PD-1 or
PD-L1
CTLA-4
?
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
CTLA-4 blockade:ipilimumab(Yervoy)
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
B7
Canonical mode of action (MoA) of ipilimumab
DendriticCell
CTLA-4
✗Inhibition
Cellule dendritique
DendriticCell
Ipilimumab
Activation
B7
CD28
CTLA-4
DendriticCell
++ + -
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
However, ipilimumab MoA is much more complex!
1 Krummel, J. Exp. Med 1995; 2 Romano, PNAS 2015; 3 Reviewed in Walker, Nat Rev Immunol 2011
2 Treg depletion by ADCC2
Treg
CTLA4
MȻ
FC-γ
APCT cell
CTLA4 CD80
p-MHCTCR
1 Blocking CTLA4 inhibitory signal1
T cell
CTLA4 CD80
APC
IDO TRP
3 Blocking IDO expression by APC3
4 Blocking inhibitory cytokines3
T cell
CTLA4 CD80
APC
TGF-𝛽
T cell
5 Blocking CD80 capture3
T cell
Soluble CTLA4
APC
T cellCD28
CD80
T cell
APC
6 Blocking CTLA4 ligands capture3
CD80
No Signal II
Endo-cytosis
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Long term benefit of CTLA-4 blockade: pooled analysis1
• Long term benefit can be in part explained by the larger number of antigen recognized
• Upon loss of an antigen by the tumors, growth can be controlled by the other ones
• Ipilimumab has been shown to increase the antigenic repertoire
• Kvistborg & al. Science Transl Med 2014
?
Checkpoint
1 Schadendorf, JCO 2015
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Tumor
antigen
Tumor cell MHC-I
CD8+
T cell
TCR
+
-
Oncogenic
event
2 PD-L1 expression resulting from oncogenic events
Tumor cell MHC-I
CD8+
T cell
TCR
STAT
IFN-γ
+
-
1 PD-L1 expression induced by IFN
Taube, Sci Transl Med 2012
Green, Blood 2010:9p24.1 amplification leads toincreased PD-L1 expression inHodgkin
Dynamic PD-L1 expression
Constitutive PD-L1 expression
PD-1 / PD-L1 axis: deciphering PD-L1 expression mechanisms
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
PD-1 / PD-L1 pathway: biological interpretation
Tumor cell Lymphocyte
+
-
=
INF-γ
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
However, the biology of PD-1/PD-L1 is much more complex!
Image courtesy
from J. Allison
PD-1 / PD-L1 Interactions
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Nivolumab, pembrolizumab,
and atezolizumab approval in NSCLC
Nivolumab approval in RCC
Nivolumab and Pembrolizumab
approval in SCCHN
Nivolumab and Atezolizumab
approval in Bladder cancer
Nivolumab and pembrolizumab
approval in Melanoma
PD-1/PD-L1 inhibition successes across multiple tumors
1000
100
10
1
0.1
0.01
n=22
Rh
abd
oid
tu
mo
r
Ewin
g sa
rco
ma
Thyr
oid
AM
L
Me
du
llob
last
om
a
Car
cin
oid
Neu
rob
last
om
a
Pro
stat
e
CLL
Low
-gra
de
glio
ma
Bre
ast
Mu
ltip
le m
yelo
ma
Pan
crea
s
Kid
ney
cle
ar c
ell
Kid
ney
pap
illar
y ce
llO
vari
an
Glio
bla
sto
ma
Cer
vica
l
DLB
CL
Hea
d a
nd
nec
k
Co
lore
ctal
Eso
ph
agea
lad
eno
carc
ino
ma
Sto
mac
h
Bla
dd
er
Lun
g A
D
Lun
g SQ
Mel
ano
ma
20 52 134 26 23 81 227 91 57 121 13 63 214 11 394 219 20 49 181 231 76 88 35 335 179 121So
mat
ic M
uta
tio
n
Freq
uen
cy (
/Mb
)1
1. Lawrence, Nature. 20135t
h ESO-E
SMO L
atin
Amer
ican
Mas
terc
lass i
n Clin
ical O
ncolo
gy
Toxicity
managment:
detailed
guidelines
exist, e.g.
ESMO
+
-
Haanen, Ann Oncol. 20175t
h ESO-E
SMO L
atin
Amer
ican
Mas
terc
lass i
n Clin
ical O
ncolo
gy
Clinical management of stage III melanoma
pT1b-T4b cN0 cM0
SLNB
Std. FU Adjuvant criteria?
US FU Adjuvant
Relapse
Adjuvant
Relapse
CLND
CLND
Negative Positive
YesNo
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Results from MSLT-2: no benefit to radical lymphadenectomy
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
MSLT-2 – Study design
Important:
• The control arm is not follow-up, but active surveillance with 4 monthly ultrasounds!
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Results from MSLT-2: no benefit to radical lymphadenectomy
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Possible stage III
melanoma
management
algorithm
pT1b-T4b cN0 cM0
SLNB
Std. FU Adjuvant criteria?
US FU Adjuvant
Relapse
Adjuvant
Relapse
CLND
CLND
Negative Positive
YesNo
Figure legends:• SLNB: sentinel lymph
node biopsy• CLND: complete lymph
node dissection• FU: follow-up
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Adjuvant treatments
100%
0%
Overall Survival
Time
Control arm
Adjuvant arm
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Risk / benefit ratio: number needed to treat
With 11% gain at 5 years, 89% of the population is exposed to treatment with no benefit and the number needed to treat is 9.1 compared to 35 for INF1
100%
0%
Overall Survival
Time
Adjuvant did notchange outcome:patient death
Adjuvant benefit: 11% @ 5y
Adjuvant did notchange outcome:patient is cured
Control arm
Adjuvant arm
1 Mocellin, Cochrane Review, 2013
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Key efficacy landmarks in the adjuvant setting of melanoma
Olivier Michielin, MD-PhD
EORTC 1325 • Pembrolizumab vs placebo, • Stage IIIA-C; RFS HR 0.57, OS HR NA
RF
S (
%)
Months
0
10
20
30
40
50
60
70
80
90
100
0 6 12 18 24 273 9 15 21
453 353 311 249 5 0399 332 291 71NIVO
453 314 252 184 2 0364 269 225 56IPI
Number of patients at risk
NIVO
IPI
66%
53%
71%
61%
Checkmate 238
• Ipilimumab 10 mg/kg vs nivolumab, • Stage IIIB-C + IV; RFS HR 0.65,
OS HR NA
Months
Nivo > Ipi
Months
Pembro > Pbo
COMBI-AD • Dabrafenib + trametinib vs placebo• Stage IIIA-C; RFS HR 0.47, OS HR 0.57
Months
D+T > Pbo
EORTC 18071 • Ipilimumab 10 mg/kg vs placebo, • Stage IIIA-C; RFS HR 0.76, OS HR 0.72
Years0 1 2 3 4 5 6 7 8
0
10
20
30
40
50
60
70
80
90
100
O N Number of patients at risk
264 475 283 217 184 161 77 13 1
323 476 261 199 154 133 65 17 0
Ipilimumab
Placebo
Ipi > Pbo
Years
Current adjuvant options shown in orange
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
EORTC 18071/CA184-029: adjuvant ipilimumab
Stratification factors:
• Stage (IIIA vs IIIB vs IIIC 1-3 positive lymph nodes vs IIIC ≥4 positive lymph nodes)
• Regions (North America, European countries and Australia)
Enrollment Period: June 2008 – July 2011
Randomized, double-blind, phase 3 study evaluating the efficacy and safety of
ipilimumab in the adjuvant setting for high-risk melanoma
Treatment up to a maximum 3 years, or until disease progression, intolerable toxicity, or
withdrawal
N=475
N=476
Week 1 Week 12 Week 24
N=951
Q3W = every 3 weeks; Q12W = every 12 weeks.
High-risk, stage III, completely resected
melanomaR
INDUCTIONIpilimumab 10 mg/kg
Q3W X4
MAINTENANCEIpilimumab 10 mg/kgQ12W up to 3 years
INDUCTIONPlaceboQ3W X4
MAINTENANCEPlacebo
Q12W up to 3 years
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
HIGHLY CONFIDENTIAL
Pa
tie
nts
a
liv
e (%
)
*Stratified by stage at
randomization
Ipilimumab Placebo
Deaths/patients 162 / 475 214 / 476
Hazard ratio (95.1%
CI)*0.72 (0.58 - 0.88)
Log-rank P value* 0.001
EORTC 18071: Overall Survival
65%
54%
5-year
difference
11%
CI = confidence interval; NR = not reached.
Years0 1 2 3 4 5 6 7 8
0
10
20
30
40
50
60
70
80
90
100
O N Number of patients at risk :
162 475 431 369 325 290 199 62 4
214 476 413 348 297 273 178 58 8
Ipilimumab
PlaceboEggermont, ESMO 2016
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Checkmate-238: adjuvant nivolumab vs. ipilimumab
Patients with high-risk, completely resected stage
IIIB/IIIC or stage IV melanoma
Enrollment period: March 30, 2015 to November 30, 2015
Follow-up
Maximum treatment duration of
1 year
NIVO 3 mg/kg IV Q2W and
IPI placebo IV Q3W for 4 doses
then Q12W from week 24
IPI 10 mg/kg IV Q3W for 4 doses
then Q12W from week 24 and
NIVO placebo IV Q2W
1:1
n = 453
n = 453
Stratified by:
1) Disease stage: IIIB/C vs IV M1a-M1b vs IV M1c
2) PD-L1 status at a 5% cutoff in tumor cells
Weber, NEJM 2017
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
RF
S (
%)
Months
0
10
20
30
40
50
60
70
80
90
100
0 6 12 18 24 273 9 15 21
453 353 311 249 5 0399 332 291 71NIVO
453 314 252 184 2 0364 269 225 56IPI
Number of patients at risk
NIVO
IPI
NIVO IPI
Events/patients 154/453 206/453
Median (95% CI) NR NR (16.6, NR)
HR (97.56% CI) 0.65 (0.51, 0.83)
Log-rank P value <0.0001
66%
53%
71%
61%
Checkmate-238: primary endpoint – relapse free survival
Weber, NEJM 20175t
h ESO-E
SMO L
atin
Amer
ican
Mas
terc
lass i
n Clin
ical O
ncolo
gy
Adjuvant BRAFi + MEKi: COMBI-AD
Key eligibility criteria• Completely resected, high-risk stage
IIIA (lymph node metastasis > 1 mm), IIIB, or IIIC cutaneous melanoma
• BRAF V600E/K mutation• Surgically free of disease ≤ 12 weeks
before randomization• ECOG performance status 0 or 1• No prior radiotherapy or systemic
therapy
R
A
N
D
O
M
I
Z
A
T
I
O
N
Stratification• BRAF mutation status (V600E, V600K)• Disease stage (IIIA, IIIB, IIIC)
1:1
Dabrafenib 150 mg BID + trametinib 2 mg
QD(n = 438)
2 matched placebos
(n = 432)
Treatment: 12 monthsa
Follow-up until end of study
• Primary endpoint: RFSd
• Secondary endpoints: OS, DMFS, FFR, safety
N = 870
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
438 413 405 392 382 373 355 336 325 299 282 276 263 257 233 202 194 147 116 110 66 52 42 19 7 2 0432 387 322 280 263 243 219 203 198 185 178 175 168 166 158 141 138 106 87 86 50 33 30 9 3 0 0
Months From RandomizationDabrafenib plus trametinibPlacebo
No. at Risk
0
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52
Pro
po
rtio
n A
live
and
Rel
apse
Fre
e1 y, 88%
2 y, 67%
3 y, 58%1 y, 56%
2 y, 44%3 y, 39%
NR, not reached.
GroupEvents,
n (%)Median
(95% CI), moHR
(95% CI)Dabrafenib+trametinib
166 (38)NR
(44.5-NR) 0.47(0.39-0.58);
P < .001Placebo 248 (57)16.6
(12.7-22.1)
P = .0000000000000153
Adjuvant BRAFi + MEKi: COMBI-AD
Hauschild, JCO 20185t
h ESO-E
SMO L
atin
Amer
ican
Mas
terc
lass i
n Clin
ical O
ncolo
gy
Comparison of stage subgroup eligibility criteria
Stage - AJCC 7th Edition (All patients NED)
Study Design IIC IIIA IIIB IIIC IV
EORTC 18071
Ipi 10 vs. placebo
✓ SN > 1mm ✓✓ no in
transit mets
EORTC 1325
Pembro vs. placebo
✓ SN > 1mm ✓✓ no in
transit mets
Checkmate 238
Ipi 10 vs. nivo ✓ ✓ ✓
ECOG 1609
Ipi 10 vs ipi 3 vs. HD INF-α2b ✓ ✓ ✓M1a-b
Olivier Michielin, MD-PhD
BRIM-8 Vem vs. placebo ✓ ✓ SN > 1mm ✓ ✓
COMBI-AD
Dabra + trame vs. placebo
✓ SN > 1mm ✓ ✓
NA, Not Available; NE, Not Estimated1 AJCC 8th Edition staging
FDA
11
.15
FDA
12
.17
FDA
04
.18
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Adjuvant in melanoma: important data are still missing!
Efficacy data
Study Design HR RFS HR DMFS HR OS
EORTC 180711
Ipi 10 mg vs. placebo
0.76 0.76 0.72
EORTC 13252
Pembro vs. placebo
0.57 0.536 NA
Checkmate 2382
Ipi 10 vs. nivo 0.65 0.737 NA
ECOG 1609
Ipi 10 vs ipi 3 vs. HD INF-α2b
1.0 NA NA
Olivier Michielin, MD-PhD
BRIM-84 Vem vs. placebo0.54 (IIC-IIIB)
0.8 (IIIC)NA NA
COMBI-AD5
Dabra + trame vs. placebo
0.47 0.51 0.57
Data not randomized head to head, should not be compared directly; NA, Not Available; NE, Not Estimated; 1 AJCC 8th Edition staging
1Eggermont, NEJM 2016; 2Eggermont NEJM 2018; 3Weber, NEJM 2017; 4Maio, Lancet Oncol 2018; 5Long, NEJM 2017;6Preliminary, Eggermont, AACR 2018; 7Exploratory; 8Faries, NEJM 2017; 9Leiter, Lancet 2016; Time in months;NA: Not Available;
Stage III patients from these trials were required to have complete lymph node dissection!
How do we integrate those results in a post MSLT-2/ DeCOG8,9 trial era?
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Overview of PFS outcome per stage subgroup
Stage - AJCC 7th Edition (All patients NED)
Study Design IIC IIIA IIIB IIIC IV
EORTC 18071
Ipi 10 mg vs. placebo
SN > 1mm, HR 0.98
HR 0.75HR 1.00, 1-3 nHR 0.48, ≥ 4 n
EORTC 1325
Pembro vs. placebo
SN > 1mm,HR 0.38
HR 0.58 HR 0.58
Checkmate 238
Ipi 10 vs. nivo HR 0.67 HR 0.65HR 0.63 M1a/b,
HR 1.0 M1c2
ECOG 1609
Ipi 10 vs ipi 3 vs. HD INF-α2b
HR NA HR NAM1a-b,HR NA
Olivier Michielin, MD-PhD
BRIM-8 Vem vs. placebo HR 0.0-NESN > 1mm,
HR 0.52HR 0.63 HR 0.8
COMBI-AD
Dabra + trame vs. placebo
SN > 1mm,HR 0.44
HR 0.50 HR 0.45
Data not randomized head to head, should not be compared directly; NA, Not Available; NE, Not Estimated; 1AJCC 8th Edition staging; 2CI 0.37-2.66!
FDA
11
.15
FDA
12
.17
FDA
04
.18
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Adjuvant immunotherapy: review of existing toxicity data
Olivier Michielin, MD-PhD
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
AEs comparison: EORTC 18071, Checkmate 067 & CA 184-169
ToxicityIpilimumab
10 mg/kg1,2, data2
Nivolumab 3 mg/kg2
Pembrolizumab200 mg3,4
Dabrafenib + trametinib5,6
All values in % All G 3-4 All G 3-4 All G 3-4 All G 3-4
Any AE 99 55 97 25 93 32 97 41
Any drug related AE 96 46 85 14 78 15 916 316
Fatigue 33 1 35 <1 37 1 47 4
Rash 29 3 20 1 16 <1 24 0
Diarrhea / colitis 46/10 10/8 24/2 2/1 19/4 1/2 33/NR 1/NR
Increased AST/ALT 13/15 4/6 6/6 <1/1 NR/NR NR/NR 14/15 4/4
Pneumonitis 2 1 1 0 3 1 - -
Hypophysitis 11 3 2 <1 2 1 - -
Adrenal disorder 3 1 1 <1 1 <1 - -
Thyroid disorder 13 1 20 1 21 <1 - -
Type I diabetes <1 <1 <1 0 1 1 - -
Olivier Michielin, MD-PhD1 Eggermont, NEJM 2016; 2 Weber, NEJM 2017; 3 Eggermont, NEJM 2018; 4 Eggermont, AACR 2018;5 Long, NEJM 2017; 6 Long, SMR 2017; NR – Not Reported5t
h ESO-E
SMO L
atin
Amer
ican
Mas
terc
lass i
n Clin
ical O
ncolo
gy
Targeted therapies for stage IV melanoma
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Preclinical programs:• Allosteric inhibitors 1
Targeted therapies in stage IV melanoma: main trials R
TK
RAS
BRAF
MEK
ERK
Melanoma cell
COMBI-d7: BRAFV600
• D+T vs. dabrafenib• RR 67%, mPFS 9.3m• HR OS/PFS: 0.63/0.75
COMBI-v6: BRAFV600
• D+T vs. vemurafenib• RR 64%, mPFS 11.4m• HR OS/PFS: 0.69/0.56
1 Ostrem, Nature 2013; 2 McArthur, Lancet Oncol 2014; 3 Hauschild, Lancet 2012; 4 Flaherty, NEJM 2012; 5 Dummer, Lancet Oncol 2017; 6 Robert NEJM 2015; 7 Long, Lancet 2015; 8 Larkin NEJM 2014; 9 Dummer, EADO 2015; 10 Wong, Molecular Cancer 2014; NA: Not Available
Co-BRIM8: BRAFV600
• V + cobi vs. V• RR 68%, mPFS 9.9m• HR OS/PFS: 0.51/0.65
COLOMBUS9: BRAFV600
• enco + bini / enco / V• RR 63%, mPFS 14.9• HR OS/PFS: NA/0.54
Preclinical & clinical:• Allosteric and ATP
competitors 10
• GDC-0994 in phase I
BREAK-33: BRAFV600
• dabrafenib vs. DTIC• RR 47%, mPFS 5.1m• HR OS/PFS: 0.61/0.30
BRIM-32: BRAFV600
• vemurafenib vs. DTIC• RR 57%, mPFS 6.9m• HR OS/PFS: 0.70/0.38
Metric4: BRAFV600
• trametinib vs DTIC• RR 22%, mPFS 4.8m• HR OS/PFS: 0.54/0.45
NEMO5: NRASmut
• binimetinib (MEK162)• RR 15%, mPFS 2.8m• HR OS/PFS: 1.0 / 0.62
• Dabrafenib• Vemurafenib• …
• Trametinib• Cobimetinib• …
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
COMBI-d: study design
Presented by Keith Flaherty, ASCO 2016
N = 947 screened
Primary Endpoint: investigator-assessed PFS
Secondary Endpoints: OS, overall response rate (ORR), duration of response, safety
N = 423
• BRAF V600E/K• Unresectable stage IIIC/IV• Treatment naive• ECOG PS 0/1• No brain metastases, unless:
Treated Stable ≥ 12 weeks
Stratification• BRAF-mutant (V600E vs K)• LDH (> ULN vs ULN)
Dabrafenib + trametinib150 mg BID + 2 mg QD
n = 211
Dabrafenib + placebo150 mg BID + placebo QD
n = 212
Pre-planned interim OS[95 events]
Primary Analysis
(PFS)[213 events]
Aug 2013
Final Analysis(OS)
[222 deaths]Jan 2015
BID, twice daily; ECOG PS, Eastern Cooperative Oncology Group performance status; LDH, lactate dehydrogenase; QD, once daily; ULN, upper limit of normal.
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
COMBI-d: 3 year survival estimates
Overall Survival
212 175 138 104 84 69 57 7 0
211 187 143 111 96 86 76 13 0
Dabrafenib + Trametinib (n = 211)
Dabrafenib + Placebo (n = 212)b
3-y OS, 44%
3-y OS, 32%
1.0
0.8
0.6
0.4
0.2
0.0
0 6 12 3018
Months From Randomization
OS
Pro
ba
bil
ity
D+Pbo
D+T
Number at risk
2-y OS, 52%
2-y OS, 43%
24 36 42 48
Progression-Free Survival
212 110 67 41 29 11 7 1 0
211 137 84 69 54 45 31 0
1.0
0.8
0.6
0.4
0.2
0.0
0
Months From Randomization
PF
S P
rob
ab
ilit
y
D+Pbo
D+T
Number at risk
6 12 3018 24 36 42 48
3-y PFS, 22%
3-y PFS, 12%
2-y PFS, 30%
2-y PFS, 16%
Dabrafenib + Trametinib (n = 211)
Dabrafenib + Placebo (n = 212)
58% of D+T patients alive at
3 years still on D+T
Presented By Keith Flaherty at 2016 ASCO Annual Meeting
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Baseline Factors Influencing OS in Combi-d, Combi-v and Ph-II
• Regression tree analysis based on:
• BMI
• Age
• LDH(N, > 1- ≤ 2x ULN, ≥ 2x ULN)
• Sex
• ECOG
• Visceral disease
• Number of sites
• Prior adjuvant immunotherapy
• 5 prognostic subgroups with very large OS differences
N=617
Adapted from:GV. Long, SMR 2015, JCO 2016K. Flaherty, ASCO 2016
N=3981Y=85%2Y=75%3Y=57%
N=2191Y=54%2Y=25%3Y=7%
Normal LDH LDH ≥ ULN
N=1491Y=60%2Y=33%3Y=9%
N=701Y=40%2Y=7%3Y=7%
LDH ≥ 2x ULN
LDH > 1-≤ 2x ULN
N=2371Y=90%2Y=75%3Y=70%
N=1611Y=76%2Y=55%3Y=38%
Sites ≥ 3Sites < 3
N=931Y=71%2Y=43%3Y=NE
N=561Y=42%2Y=19%3Y=16%
ECOG ≥ 1ECOG = 1
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
3-year OS data from COLOMBUS
Overa
ll S
urv
ival, %
Time (months)
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 21 24 27 36 39 42 4815 18 30 33 45
ENCO300
VEM
Censored patients
Median OS in months (95% CI)
ENCO300 VEM
23.5 (19.6–33.6) 16.9 (14.0–24.5)
HR (95% CI), 0.76 (0.58–0.98)
Nominal 2-sided P=0.033
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Preclinical programs:• Allosteric inhibitors 1
Targeted therapies in stage IV melanoma: main trials R
TK
RAS
BRAF
MEK
ERK
Melanoma cell
COMBI-d7: BRAFV600
• D+T vs. dabrafenib• RR 67%, mPFS 9.3m• HR OS/PFS: 0.63/0.75
COMBI-v6: BRAFV600
• D+T vs. vemurafenib• RR 64%, mPFS 11.4m• HR OS/PFS: 0.69/0.56
1 Ostrem, Nature 2013; 2 McArthur, Lancet Oncol 2014; 3 Hauschild, Lancet 2012; 4 Flaherty, NEJM 2012; 5 Dummer, Lancet Oncol 2017; 6 Robert NEJM 2015; 7 Long, Lancet 2015; 8 Larkin NEJM 2014; 9 Dummer, EADO 2015; 10 Wong, Molecular Cancer 2014; NA: Not Available
Co-BRIM8: BRAFV600
• V + cobi vs. V• RR 68%, mPFS 9.9m• HR OS/PFS: 0.51/0.65
COLOMBUS9: BRAFV600
• enco + bini / enco / V• RR 63%, mPFS 14.9• HR OS/PFS: NA/0.54
Preclinical & clinical:• Allosteric and ATP
competitors 10
• GDC-0994 in phase I
BREAK-33: BRAFV600
• dabrafenib vs. DTIC• RR 47%, mPFS 5.1m• HR OS/PFS: 0.61/0.30
BRIM-32: BRAFV600
• vemurafenib vs. DTIC• RR 57%, mPFS 6.9m• HR OS/PFS: 0.70/0.38
Metric4: BRAFV600
• trametinib vs DTIC• RR 22%, mPFS 4.8m• HR OS/PFS: 0.54/0.45
NEMO5: NRASmut
• binimetinib (MEK162)• RR 15%, mPFS 2.8m• HR OS/PFS: 1.0 / 0.62
• Dabrafenib• Vemurafenib• …
• Trametinib• Cobimetinib• …
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
3
Study Design and Objectives
COMBO450=encorafenib 450 mg QD + binimetinib 45 mg BID; ECOG PS=Eastern Cooperative Oncology Group performance status; OS=overall survival; PFS=progression-free survival; R=randomization; VEM=vemurafenib 960 mg BID.
*Amendment requested by FDA.†Included in hierarchical testing approach.‡Median follow-up of patients assessed using reverse Kaplan-Meier approach (i.e. median potential follow-up).
Efficacy update with additional follow-up of 18 months:OS:
Secondary endpoint†
Planned after 232 events in the COMBO450 and VEM groups combined
Median duration of follow-up‡: 36.8 months
PFS:
Primary endpoint
Median duration of follow-up‡: 32.1 months
R
3:1
(N=344)
COMBO300Encorafenib 300 mg QD +
Binimetinib 45 mg BID (n=258)
ENCO300Encorafenib 300 mg QD (n=86)
COLUMBUS
Part 2* Untreated or progressed
on/after prior first-line
immunotherapy
BRAFV600E and/or BRAFV600K
ECOG PS 0–1
R
1:1:1
(N=577)
COMBO450Encorafenib 450 mg QD +
Binimetinib 45 mg BID (n=192)
VEM Vemurafenib 960 mg BID (n=191)
ENCO300Encorafenib 300 mg QD (n=194)
COLUMBUS Part 1
Reinhard Dummer
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Median OS in months (95% CI)
COMBO450 VEM
33.6 (24.4–39.2) 16.9 (14.0–24.5)
HR (95% CI), 0.61 (0.47–0.79)
Nominal 2-sided P<0.0001
Overall Survival: COMBO450 vs VEM
5
1
0
10
20
30
40
50
60
70
80
90
100O
ve
rall S
urv
iva
l, %
COMBO450
VEM
Patients at risk Time (months)
0 3 6 9 12 15 18 21 24 27 30 33 36 39 42 45 48
192 188 182 166 144 132 124 115 108 102 95 82 57 30 9 1 0
191 184 166 140 115 100 89 83 77 71 62 56 30 19 8 1 0
COMBO450
VEM
Censored patients
Reinhard Dummer
COMBO450=encorafenib 450 mg QD + binimetinib 45 mg BID; HR=hazard ratio; OS=overall survival; VEM=vemurafenib 960 mg BID.
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
5
2
Ove
rall S
urv
iva
l, %
Time (months)
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 21 24 27 36 39 42 4815 18 30 33 45
ENCO300
VEM
Censored patients
194 181 168 147 133 117 109 94 86 83 79 59
191 184 166 140 115 100 89 83 77 71 62 56
ENCO300
VEM
Patients at risk
40 24 5 0 0
30 19 8 1 0
Overall Survival: ENCO300 vs VEM
ENCO300=encorafenib 300 mg QD; HR=hazard ratio; OS=overall survival; VEM=vemurafenib 960 mg BID.
Median OS in months (95% CI)
ENCO300 VEM
23.5 (19.6–33.6) 16.9 (14.0–24.5)
HR (95% CI), 0.76 (0.58–0.98)
Nominal 2-sided P=0.033
Reinhard Dummer
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Immuno-therapies for stage IV melanoma
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Checkmate-067: PD-1 blockade as a backbone for combinations1
NIVO
37%32% 31%
Months
PFS
(%
)
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18 573024 362721 45 5133 39 5442 48
aDescriptive analysis
1 Hodi, ESMO 20182 Sharma, Cell 2017
Innate resistance2:• Lack of antigenic mutations, MHC loss• Constitutive expression of PD-L1• Increase in T-regs, MDSCs, TAMs• Immunosuppressive cytokines: IL-10, TGF-𝛽• Upregulation of checkpoints: LAG-3, TIM-3• …
Acquired resistance2:• Loss of target antigen• 𝛽2m (light chain of MHC)
mutations• Escape mutation in INF
pathway (e.g. JAK/STAT)
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Rational combination or rational sequencing?
Possible strategies for I-O sequences or combos
2nd I-O agentPD-1 PD
1 Sequencing
PD-1 PD-1PD
2nd I-O agent
2 Combo at relapse
PD-1
2nd I-O agent
3 Combo at start
Depending on the type of resistance mechanism, various strategies might be more appropriate to obtain maximal global PFS. Biomarkers are required to guide strategies.
Resistance mechanism
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Rational selection of PD-1 based combo at start
PD-1 + CTLA-4 inhibitor data: Checkmate-067
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Unresectable or
metatastic melanoma
• Previously untreated
• 945 patients
Treat until
progression
or
unacceptable
toxicity
NIVO 3 mg/kg Q2W +IPI-matched placebo
NIVO 1 mg/kg + IPI 3 mg/kg Q3W for
4 doses then NIVO 3 mg/kg Q2W
IPI 3 mg/kg Q3W for 4 doses +
NIVO-matched placebo
Stratify by:
• BRAF status
• AJCC M stage
• Tumor PD-L1 expression <5% versus ≥5%
n = 314
n = 316
n = 315
4-year follow up of a randomized, double-blind,
phase 3 study to compare NIVO+IPI or NIVO
alone with IPI alonea
Database lock: May 10, 2018; minimum follow-up of 48 months for all patients
CheckMate 067: Study Design
57
Co-primary endpointsa were PFS and OS in
the NIVO-containing arms versus IPI alone
R
1:1:1
aThe study was not powered for a comparison between NIVO+IPI and NIVO
Hodi, ESMO 2018; Hodi, Lancet Oncol 2018
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Response to Treatment
NIVO+IPI
(n = 314)
NIVO
(n = 316)
IPI
(n = 315)
ORR, % (95% CI) 58.3 (52.6, 63.8) 44.6 (39.1, 50.3) 19.0 (14.9, 23.8)
Best overall response, %
Complete response 21.3 17.7 5.1
Partial response 36.9 26.9 14.0
Stable disease 12.1 9.5 21.6
Progressive disease 23.6 38.3 50.5
Unknown 6.1 7.6 8.9
Median duration of response, months (95% CI)
50.1 (44.0, NR) NR (45.7, NR) 14.4 (8.3, NR)
58NR = not reached Hodi, ESMO 2018; Hodi Lancet Oncol 2018
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
NIVO+IPI (n = 314) NIVO (n = 316) IPI (n = 315)
Median OS, mo (95% CI)NR
(38.2, NR)36.9
(28.3, NR)19.9
(16.9, 24.6)
HR (95% CI) versus IPI0.54
(0.44, 0.67)0.65
(0.53, 0.79)–
HR (95% CI) versus NIVOa 0.84 (0.67, 1.05)
– –
Overall Survival
59
Months
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18 573024 362721 45 5133 39 5442 48
Patients at risk:
0IPI 253285315 227 203 181 163 148 135 128 113 107 99 94 93 90 86 50 11
0265292314NIVO+IPI 247 226 221 209 200 198 192 186 180 178 171 166 160 154 96 13
NIVO 0266292316 245 231 214 201 191 181 175 171 164 158 150 144 140 135 85 18
64%58%
53%
59%
51%46%45%
34%30%
aDescriptive analysis
OS
(%
)
NIVO+IPI
NIVO
IPI
Hodi, ESMO 2018; Hodi Lancet Oncol 2018
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Current IO developments in 1L stage IV melanoma
NIVO+IPI
NIVO
IPI
Patients at risk:
0IPI 78136315
0175218314NIVO+IPI
NIVO 0151177316
58
155
132
46
136
120
42
131
112
34
124
106
32
117
103
31
110
97
29
104
88
28
101
84
26
95
79
19
93
77
18
89
75
16
88
72
16
81
66
11
53
50
7
19
18
1
3
0
41%
37%
12%
39%
32%
10%
37%
31%
9%
Months
PFS
(%
)
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18 573024 362721 45 5133 39 5442 48
Checkmate 067:Still a highunmet medicalneed!
Hodi, ESMO 2018
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
OS by Tumor PD-L1 Expression, 5% Cutoff
PD-L1 Expression Level <5% PD-L1 Expression Level ≥5%
NIVO+IPI NIVO IPI
Median, mo
(95% CI)
NR
(32.7, NR)
35.9
(23.1, NR)
18.4
(13.7, 22.5)
HR (95% CI)
versus IPI
0.54
(0.42, 0.69)
0.64
(0.50, 0.82)–
HR (95% CI)
versus NIVOa
0.83
(0.64, 1.09) – –
NIVO+IPI NIVO IPI
Median, mo
(95% CI)
NR
(39.1, NR)
NR
(33.6, NR)
28.9
(18.1, 44.2)
HR (95% CI)
versus IPI
0.56
(0.35, 0.90)
0.65
(0.42, 0.99)–
HR (95% CI)
versus NIVOa
0.86
(0.53, 1.41) – –
56%
50%
30%
52%
45%
28%
65%
59%
45%
61%
54%
36%NIVO+IPI
NIVO
IPI
MonthsPatients at risk:
0667275 64 60 55 46 43 40 39 34 34 31 28 28 26 24 13 3IPI
0566368 55 52 50 45 45 45 44 43 43 43 42 41 41 38 24 3NIVO+IPI
0767980 74 69 64 61 58 57 54 53 50 47 46 43 42 41 24 6NIVO
OS
(%
)
0
20
40
60
80
100
0 3 6 9 12 15 18 573024 362721 45 5133 39 5442 48
OS
(%
)
0
20
40
60
80
100
0 3 6 9 12 15 18 573024 362721 45 5133 39 5442 48
0158179202
0178194210
0169189208
140
163
151
124
146
144
107
144
133
99
139
123
89
131
118
80
130
112
77
127
110
69
123
108
64
118
104
59
116
102
58
111
95
57
107
92
56
103
89
55
100
86
31
62
56
6
9
10
NIVO+IPI
NIVO
IPI
MonthsPatients at risk:
IPI
NIVO+IPI
NIVO
aDescriptive analysis
63%
55%
41%
68%
71%
54%
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Imperfect current biomarkers for patient selection1
• No good predictive biomarkers have been identified that allow to strongly separate the patient populations
• PD-L1 and BRAF allow to select population with higher benefit, but the biomarker negative population still derives benefit
False positive rate
Tru
e p
osi
tive
rat
e
ROC curves confirm the poor performance of PD-L1 to guide patient selection: Fig. S4
1Wolchok, NEJM 20175t
h ESO-E
SMO L
atin
Amer
ican
Mas
terc
lass i
n Clin
ical O
ncolo
gy
PD-1 based combo at start
Failure of IDO combined with Pembrolizumab
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Tumorcell
INFγ
PD-1 / PD-L1 and IDO in the T cell inflamed phenotype
• INFγ release triggers both PD-L1 and IDO in tumor cells and in the µ-env.
• Could such co-occurrence be an argument to initiate combo at start?
STAT
INFγ
TCR
T cell
MHC
PD-1
Tryptophan
Kynurenine
+
-
PD-L1
- IDO
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Study Design: Phase III Randomized Controlled Trial
6
5
Key Eligibility Criteria
• Unresectable stage III or IV melanoma,
advanced/metastatic disease
– Patients with BRAF mutation could have
received prior BRAF/MEK therapy
– Prior anti-CTLA-4 or interferon in adjuvant
setting permitted
• ECOG performance status 0–1
• No active CNS metastases
Stratification• PD-L1 status (positivea vs negative)
• BRAF mutation status
– Wild type
– Mutant with prior BRAF-directed therapy
– Mutant without prior BRAF-directed therapy
Epacadostat 100 mg PO BID
+
Pembrolizumab 200 mg IV
Q3W
n=354
Placebo
+
Pembrolizumab 200 mg IV
Q3W
n=352
• Primary endpoints: PFS (RECIST v1.1) and OS
• Secondary endpoints: ORR (RECIST v1.1),
DOR, safety
BID, twice daily; DOR, duration of response; ECOG, Eastern Cooperative Oncology Group; ORR, objective response rate; OS, overall survival; PD-L1, programmed death-ligand 1; PFS, progression-free survival;
Q3W, every 3 weeks; RECIST, Response Evaluation Criteria In Solid Tumors.a1% staining in tumor and adjacent immune cells as assessed by IHC (22C3 antibody).
Georgina V. Long6
5
N=706
R 1:1
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Progression-Free Survival (RECIST v1.1, BICR)
BICR, blinded independent central review; CI, confidence interval; E, epacadostat; HR, hazard ratio; P, pembrolizumab; PFS, progression-free survival; RECIST, Response Evaluation Criteria In Solid Tumors.
PFS defined as time from randomization to disease progression or death, whichever occurred first.
Georgina V. Long6
6
100
90
80
70
60
50
40
30
20
10
0
Pro
gre
ssio
n-F
ree
Su
rviv
al (%
)
0 2 4 6 8 10 12 14 16 18
Time, months
354
352
309
304
181
181
155
151
137
132
114
109
57
65
25
28
5
7
0
0
E + P
Placebo + P
Number at risk
36.9%36.6%
45.8%45.8%
E + P
Placebo + P
HR (95% CI): 1.00 (0.831.21)
P = 0.517
Events,
n (%)
Median PFS, months
(95% CI)
E + P 218 (61.6) 4.7 (2.96.8)
Placebo + P 219 (62.2) 4.9 (2.96.8)
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Overall Survival
CI, confidence interval; E, epacadostat; HR, hazard ratio; NR, not reached; OS, overall survival; P, pembrolizumab.
Georgina V. Long6
7
100
90
80
70
60
50
40
30
20
10
0
Ove
rall
Su
rviv
al (%
)
0 2 4 6 8 10 12 14 16 18
Time, months
354
352
340
342
322
323
290
304
274
285
263
263
183
186
96
115
42
43
5
2
E + P
Placebo + P
Number at risk
74.4%74.1%
84.1%87.2%
E + P
Placebo + P
HR (95% CI): 1.13 (0.86–1.49)
P = 0.807
Events,
n (%)
Median OS,
months
(95% CI)
E + P 106 (29.9) NR (NR, NR)
Placebo + P 98 (27.8) NR (NR, NR)
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Can one explain KN-252 failure?
• Patient population was not selected on baseline biomarkers
• IDO biology could reveal more complex than anticipated. There are partly overlapping enzymes for amino-acid catabolism1
Unresectablestage III / IV melanoma
R
Keytruda + Epacadostat
Keytruda + Placebo
ECHO-301/KEYNOTE-252 Design:
Co-Primary Endpoints:• PFS – failed• OS – likely to fail
• Most IDO-1 inhibitors are not potent inhibitors of TDO
• Epacadostat inhibits TDO very weakly with EC50 > 10 µMol
1 Molinier-Frenkel, FEBS Letter 2017
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Rational selection of PD-1 based combo at start
PD-1 + LAG-3
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
LAG-3 and T cell exhaustion
1 Marraco, Front Immunol 2015ESMO IO 2018 | Geneva | December 14th 2018
• The systems biology of T cell exhaustion is now better understood. For a review, see 1
• Antigen persistence together with continuous checkpoint engagement leads to the exhausted phenotype
• Key checkpoints involved:
• PD-1, CTLA, BTLA, TIM3, 2BA, LAG3, …
I-Rlevels
Immune response
TCR
T cellCTLA-4
LAG-3
KLRG1
PD-1
TIM3
2B4
BTLA
TCR
T cell
TIM3
BTLA
TCR
T cellCTLA-4
LAG-3
PD-1
TIM3
2B4
BTLA
⛔️
Naive
Effector(activated)
Exhausted
TCR
T cell
KLRG1
2B4
BTLA
Memory
Persistent Ag + IR:IRL
Ag cleared
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Pink: PD-L1 ≥ 1% Blue: PD-L1 < 1% Gray: PD-L1 unknown
100
80
60
40
20
0
-20
-40
-60
-100
-80
100
80
60
40
20
0
-20
-40
-60
-100
-80
100
80
60
40
20
0
-20
-40
-60
-100
-80
45% with tumor
reduction
24% with tumor
reduction
13% with tumor
reduction
Be
st
pe
rce
nt
ch
an
ge
in s
um
of
targ
et
les
ion
dia
me
ters
fro
m b
as
eli
ne
a,b
aSix patients with clinical progression prior to their first scan and 1 with PD due to a new symptomatic brain metastasis prior to getting full
scans were not included. bOne patient with best change from baseline > 30% had a best response of SD.
LAG-3 ≥ 1%n = 29
LAG-3 < 1%n = 17
LAG-3 Unknownn = 8
PD-1 + LAG-3 combo in patient relapsing on PD-1 single agent1
1 Ascierto, ESMO 2017
Towards a rational PD-1 based combo selection based LAG-3?
(ORR 18%vs. 11.5 for total pop.)
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Towards complex, multidimensional predictive biomarkers
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Searching for better biomarkers for stage IV patient selection
?
Wolchok, NEJM 2017
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Checkmate-038 prospective biomarker study1
1Riaz, Cell 2017 (in print)
• 68 advanced melanoma patients
• Multidimensional biomarker analysis at baseline and on treatment reveals molecular actions of anti-PD-1 therapy
a) Anti-PD-1 therapy induces changes in the mutational burden of tumors
b) Distinct changes in gene expression programs associate with clinical response
c) Shifts in the TCR repertoire occur following immune checkpoint blockade
• This study is one example of the level of information that need to be integrated for complex biomarker research
• Many more to come!
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
2nd target selection for PD-1 based combos:
APCT cell
CTLA4 CD80Ag
Tumor
T cell
PD-1 PD-L1
αPD-1 +
Non Immune Microenvironment• Anti-angiogenesis (VEGF)• Depleting CAFs (NOX4, …)
Tumor cell intrinsic• Driver oncogenes (BRAFi, MEKi)• AXL, Wnt, TGF𝛽 inhibition• Radiotherapy
Priming• Oncolytic viruses (TVEC)• Other IT agents
(TLR agonists)• Checkpoints (CTLA-4, …)• Radiotherapy
CD8+ T cell intrinsic• T cell proliferation (IL-2)• Exhaustion checkpoints
(LAG-3, TIM-3, VISTA, …)• Suppressive metabolites
(IDO, TDO, Arginase, A2a)• Epigenetic reprogramming
(HDACi)
T cell tumor site homing• Tumor endothelium modulation
(Endothelin receptors)• Anti-angiogenesis (VEGF)• Fas/Fas-L inhibition
T Reg
Non CD8 Immune Microenvironment(T, MDSC, Neutrophils)• T reg depletion (CTLA-4)• MDSC depletion• M2 Macrophage
depletion (CSF-1R)• Neutrophils
(neutrophil trap)
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Conclusion and outlook
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Conclusion: modern treatment of melanoma, a whole new world!
• In less than 10 years, the treatment of melanoma has been completely rewritten
• Two strategies have provided unprecedented overall survival benefits
• Targeted and immuno-therapies
• Our challenge for the years to come is
• to optimally combine and/or sequence them
• to define predictive biomarkers for precise patient selection and maximal benefit, i.e. a cure!
BRAFi1,2
Early clinical benefit
CTLA-4 blockade3
Late clinical benefit
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Possible treatment algorithm for stage IV melanoma1
stLi
ne
2n
dLi
ne
3rd
Lin
e
BRAF WT BRAF Mutated NRAS Mutated
Trial
Trial
Trial
PD-1
CTLA-4
Chemo
PD-1/CLTA-4
BRAFi/MEKi
Chemo
BRAFi/MEKi
PD-1/CLTA-4
Chemo
PD-1/CLTA-4
MEKi
Chemo
PD-1
BRAFi/MEKi
CTLA-4
Chemo
PD-1
CTLA-4
MEKi
Chemo
PD-1/CLTA-4
Chemo
I-O Re-challenge
? ? ? ?
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Thank you for your attention!
March 25th 2019, Berlin, GermanyESMO/ESO Masterclass in Clinical Oncology
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Counteracting BRAF/MEKiresistance mechanisms
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Strategies to overcome BRAFi + MEKi resistance
• Combination based on molecular escape
• E.g: LOGIC-2 trialCombo
Rational 3rd TKI
PD
Combo
Combo Combo Combo
Combo PD
Combo
• Sequential regimens
• Ongoing
• Treatment beyond PD
• +/- local therapy
• Engaging the immune system
• Synergistic with combo?Combo Combo
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Example of rational combination trial: LOGIC-2
• Strategies for rational combination at PD based on molecular escapes are now emerging
• An example is the LOGIC-2 trial: NCT02159066
• Primary endpoint: ORR
• Secondary endpoints: PFS, OS, …
• Recruitment has been completed
• First data expected for 12.2017!
Binimetinib (MEK)
PD
Encorafenib (BRAF)140 stage IVBRAFV600
patients
LEE011
BGJ398
BKM120
INC280
CDK4/6 inhibitor or
FGFR inhibitor or
PI3K inhibitor or
MET inhibitor or
Mandatory baselineand PD biopsies
Binimetinib (MEK)Encorafenib (BRAF)
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Adding CDK4i to the BRAFi and MEKi backbone?
• Results from a phase I/II trial testing the triple combination of BRAF, MEK and CDK4 presented at ASCO1
• 42 patient in the phase II, BRAFi naïve were given
• encorafenib 200 mg (BRAF)
• binimetinib 45 mg (MEK)
• ribociclib 600 mg (CDK4) (3w out of 4)
• ORR was 52% compared to 63% in combo arm of COLOMBUS
• mPFS 9.2 m compared to 14.9 m in combo arm of COLOMBUS
• Explanation?
• Lower dosage of BRAF and / or MEK?
• High discontinuation rate?
• PK/PD with triple combination?1Abstract 9518, Ascierto, ASCO 2017, Discussed by G. McArthur
Need more TR correlative data
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Strategies to overcome BRAFi + MEKi resistance
• Combination based on molecular escape
• E.g: LOGIC-2 trialCombo
Rational 3rd TKI
PD
Combo
Combo Combo Combo
Combo PD
Combo
• Sequential regimens
• Ongoing
• Treatment beyond PD
• +/- local therapy
• Engaging the immune system
• Synergistic with combo?Combo Combo
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Rechallenge: prospective trial!
Schreuer, Lancet Oncol 2017
• BRAF + MEKi rechallenge, prospective phase II trial
• Patients having failed BRAFi or BRAF + MEKi with a drug free interval of 3+ months
• RR 32% (8/25)
• Baseline BRAF v600Mut cfDNA was not associated with clinical benefit, but reponders had a statistically more important cfDNA decline at W2
• Could resistance mechanisms predict clinical benefits in rechallenge?
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
3
Study Design and Objectives
Untreated or progressed
on/after prior first-line
immunotherapy
BRAFV600E and/or BRAFV600K
ECOG PS 0–1
R
1:1:1
(N=577)
COMBO450Encorafenib 450 mg QD +
Binimetinib 45 mg BID (n=192)
VEM Vemurafenib 960 mg BID (n=191)
ENCO300Encorafenib 300 mg QD (n=194)
COLUMBUS Part 1
Time (mo)
Pro
gre
ss
ion
-Fre
e S
urv
ival
(%)
100
0 4 8 12 16 20 24 28
0
20
40
60
80
COMBO450VEM
Median PFS in months (95% CI)
COMBO450 VEM
14.9 (11.0–18.5) 7.3 (5.6–8.2)
HR (95% CI), 0.54 (0.41–0.71)
P<0.0001
Dummer, R et al. Lancet Oncol. 2018; 19(5):603-615
Efficacy update with additional follow-up of 18 months:OS:
Secondary endpoint†
Planned after 232 events in the COMBO450 and VEM groups combined
Median duration of follow-up‡: 36.8 months
PFS:
Primary endpoint
Median duration of follow-up‡: 32.1 months
Reinhard Dummer
COMBO450=encorafenib 450 mg QD + binimetinib 45 mg BID; ECOG PS=Eastern Cooperative Oncology Group performance status; OS=overall survival; PFS=progression-free survival; R=randomization; VEM=vemurafenib 960 mg BID.
*Amendment requested by FDA.†Included in hierarchical testing approach.‡Median follow-up of patients assessed using reverse Kaplan-Meier approach (i.e. median potential follow-up).
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Therapeutic opportunities:towards a rational selection of PD-1 based combos?
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
A large number of checkpoints are targeted in clinical trials
T cell
TCR
CTLA-4
PD-1
TIM-3
BTLA
VISTA
LAG-3
CD28
OX40
GITR
CD137
HVEM
CD27
InhibitoryCheckpoints
ActivatingCheckpoints
Blocking
MABs
Agonistic
MABs
Adapted from Mellman, Nature 2011
Combinatorial issue!• 12*11=132
possible doublets☞We need to guidetrial development!
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Towards a rational selection of PD-1 based combo?
• Either baseline or on treatment biopsies can help guide decision
• All combos are being tested within clinical trials
• Complex biomarker will help optimal patient selection
PD-L1 positivity ≥ 1% αPD-1 alone?
PD-L1 positivity < 1% αPD-1 + αCTLA-4?
High content in TAM αPD-1 + αCSF1R?
High IDO expression αPD-1 + IDOi?
T cell exhaustion, LAG3+? αPD-1 + αLAG3?
T cell exhaustion, TIM3+? αPD-1 + αTIM3?
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Selected PD-1-based checkpoint combos tested for melanomaIndication,
clinical phaseCompound Checkpoint target PD-1 inhibitor ClinicalTrial.gov ID
Melanoma, P I-II Ipilimumab CTLA-4 Inh
ibito
ry Inh
ibito
ry Inh
ibito
ry
Pembrolizumab NCT02089685
Solid tumors, P I-II BMS-986218 CTLA-4 Nivolumab NCT03110107
Solid tumors, P I-II BMS-986016 LAG-3 Nivolumab NCT01968109
Solid tumors, P I-II LAG525 LAG-3 PDR001 NCT02460224
Solid tumors, P I-II Lirilumab KIR Nivolumab NCT01714739
Solid tumors, P I-II BMS-986207 TIGIT Nivolumab NCT02913313
Solid tumors, P I MK-7684-001 TIGIT Pembrolizumab NCT02964013
Solid tumors, P I Enoblituzumab B7-H3 Pembrolizumab NCT02475213
Melanoma, P III Epacadostat IDO Pembrolizumab NCT02752074
Solid tumors, P I-II BMS-986205 IDO Nivolumab NCT02658890
Solid tumors, P I-II Urelumab CD137 (4-1BB) Activatin
gNivolumab NCT02253992
Solid tumors, P I-II BMS-986156 GITR Nivolumab NCT02598960
Solid tumors, P I-II BMS-986178 OX40 Nivolumab NCT02737475
Solid tumors, P I-II Varlilumab CD27 Nivolumab NCT02335918
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Rational combination or rational sequencing?
Possible strategies for I-O sequences or combos
2nd I-O agentPD-1 PD
1 Sequencing
PD-1 PD-1PD
2nd I-O agent
2 Combo at relapse
PD-1
2nd I-O agent
3 Combo at start
…
Depending on the type of resistance mechanism, one strategy or the other might be more appropriate to obtain maximal global PFS. Biomarkers are required to guide such strategies.
Resistance mechanism
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
OS in Patients With BRAF Wild-type and Mutant Tumors
BRAF Wild-type BRAF Mutant
NIVO+IPI NIVO IPI
Median, mo
(95% CI)
39.1
(27.5, NR)
34.4
(24.1, NR)
18.5
(14.1, 22.7)
HR (95% CI)
versus IPI
0.60
(0.47, 0.77)
0.65
(0.51, 0.83)–
HR (95% CI)
versus NIVOa
0.92
(0.71, 1.20)– –
NIVO+IPI NIVO IPI
Median, mo
(95% CI)NR
45.5
(26.4, NR)
24.6
(17.9, 31.0)
HR (95% CI)
versus IPI
0.45
(0.30, 0.67)
0.64
(0.44, 0.93)–
HR (95% CI)
versus NIVOa
0.70
(0.46, 1.07)– –
MonthsPatients at risk:
IPI
NIVO+IPI
NIVO
0165194215
0169193211
0180199218
146
156
164
132
143
156
117
141
145
105
132
134
95
126
127
86
125
124
81
119
119
72
115
116
70
109
111
64
108
106
62
102
102
61
99
98
58
98
96
57
94
93
33
54
56
9
6
12
53%
49%
32%
49%
45%
28%
NIVO+IPI
NIVO
IPI
68%
56%
37%
62%
50%
33%
MonthsPatients at risk:
IPI
NIVO+IPI
NIVO
08891100
09699103
0869398
81
91
81
71
83
75
64
80
69
58
77
67
53
74
64
49
73
57
47
73
56
41
71
55
37
71
53
35
70
52
32
69
48
32
67
46
32
62
44
29
60
42
17
42
29
2
7
6
NIVO+IPI
NIVO
IPI
OS
(%
)
0
20
40
60
80
100
0 3 6 9 12 15 18 573024 362721 45 5133 39 5442 48
OS
(%
)
0
20
40
60
80
100
0 3 6 9 12 15 18 573024 362721 45 5133 39 5442 48
aDescriptive analysis 5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology
Moving towards phase II/III: CA224-047
Phase II
Adapted from Ascierto, ASCO 2018ESMO IO 2018 | Geneva | December 14th 2018
Unresectable or metastatic melanoma• Previously untreated• Tissue available for
LAG-3, PD-L1, TMB assessment
Stratified by:• LAG-3 status• PD-L1 status• BRAF status• AJCC M-stage
R
Relatlimab + Nivolumab160/480 mg IV Q4W
Nivolumab480 mg IV Q4W
• Phase II primary endpoint: ORR assessed by a BICR
• Phase II secondary endpoint: ORR, DOR, DCR, PFS rates, and 1- and 2-year OS rates according LAG-3 and
PD-L1 status, safety and tolerability
• Phase III primary endpoint: PFS
• Phase III secondary endpoint: ORR, OS
Phase III
Relatlimab + Nivolumab160/480 mg IV Q4W
Nivolumab480 mg IV Q4W
InterimAnalysis R
400 pts 300 pts
Clinicaltrial.gov identifier NCT03470922
5th
ESO-ESM
O Lat
in Am
erica
n M
aste
rclas
s in
Clinica
l Onc
ology