Download ASCO 2015

51st ASCO Annual Meeting, Chicago
Roche Analyst Event
Sunday, 31 May 2015
Agenda
Welcome
Karl Mahler, Head of Investor Relations
Oncology strategy and outlook
Daniel O’Day, Chief Operating Officer, Roche Pharmaceuticals
Working along the cancer immunity-cycle: Strategies and new agents
Ira Mellman, gRED: Ph.D., Vice President, Cancer Immunology, Genentech
William Pao, pRED: M.D., Ph.D., Global Head Oncology Discovery and Translational Area, Roche
ASCO 2015 Roche highlights: Setting new standards, developing combinations
Sandra Horning, M.D., Chief Medical Officer and Head Global Product Development
Growing importance of molecular information in cancer immunotherapy
Garret Hampton, Ph.D., Vice President, Oncology Biomarker Development and Companion
Diagnostics
Summary and Q&A
Karl Mahler, Head of Investor Relations
2
Oncology strategy and outlook
Daniel O’Day
Chief Operating Officer, Roche Pharmaceuticals
3
Roche oncology: Continued sales growth
A portfolio of differentiated medicines
CHF m
24,000
22,000
20,000
18,000
16,000
14,000
12,000
10,000
8,000
6,000
4,000
10 medicines
16 tumor types
42 indications
2,000
0
1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
Sales at 2014 average exchange rates
4
Oncology continues to become more complex
Complexity
Science, clinical practice evolving with understanding
of disease biology
Unknown
>1 mutated gene 3%
MEK1 <1%
NRAS 1%
MET 1%
PIK3CA
BRAF 1%
2% 3%
HER2
EGFR (other)
4%
EGFR
No oncogenic
driver detected
35%
ALK
8%
KRAS
25%
EGFR
17%
PD-L1 positive
(illustrative)
KRAS
2004
2014
Classification of lung adenocarcinomas
Pao & Girard. Lancet Oncol 2011; Johnson, et al. ASCO 2013
Today
5
Roche’s oncology strategy has remained constant
Today’s presentations highlight the results of our
strategy across key programs
Follow the science
Personalized
healthcare
•  Bringing our best minds
together
•  Science driven programs
•  Differentiation via
biomarkers
•  Experts integrated in R&D
Strong portfolio &
development
•  Comprehensive pipeline
and active partnering
•  Smart trial design
Higher certainty of patient benefit: Better for patients and for reimbursement
6
Strategy in action - Immunotherapy
Perspectives in key franchises
Summary
7
Strategy into action: Learning loop
Comprehensive and scientifically driven strategies –
crafted by our best minds
Research
Development
(gRED, pRED, Chugai &
external)
(early and late)
We bring our best
scientists & clinicians
together for:
•  Ideation
Immunotherapy &
Combinations
Partnering
(Genentech and Roche)
•  Decision making
•  Quick & efficient pursuit of
promising ideas
Clinical Insights
& Biomarkers
(internal and external)
...and we dedicate budget for
rapid proof of concept
studies (Ph1)
8
Atezolizumab (aPD-L1, MPDL3280A) strategy
Biomarker selection and combinations
Our vision: Bringing the potential for transformative benefit to a broad patient
population either in mono therapy or combinations
Diagnostic selection to identify
patients most likely to benefit from
atezolizumab as a monotherapy
PD-L1+
(tumor – infiltrating
immune cells)
Improve survival using well-tolerated
combinations with existing and
emerging therapies
PD-L1+
(tumor cells)
9
Atezolizumab in 2/3L NSCLC (POPLAR)
Clinical outcomes correlate with PD-L1 expression
Overall Survival
ITT
16%*
TC2/3 or IC2/3
37%*
TC1/2/3 or IC1/2/3
67%*
TC0 and IC0
32%*
0
0.98
0.77
n=287
TC3 or IC3
0.57
0.46
0.70
0.56
0.87
0.63
1.12
1.17
1.87
1.93
0.5
1
HRa
In favor of atezolizumab
= Unstratified HR; * = eligible patient population
Spira A. et al, ASCO 2015
a
Progression Free Survival
1.5 0
In favor of docetaxel
0.5
1
HRa
In favor of atezolizumab
1.5
In favor of docetaxel
10
Atezolizumab strategy: Combinations
Scientifically driven strategy that combines with chemo,
targeted therapy, and other immunotherapy
T cell Trafficking
pRED
gRED
Priming & activation
Eliminate M2
Macrophages
Anti-CSF-1R
Both
T cell infiltration
Costimulatory
antibodies
Anti-OX40 (agonist)
T cell bispecifics
Anti-CEA/CD3 TCB
Antigen presentation
APC stimulators
Anti-CD40 (agonist)
Cancer T cell recognition
Antigen release
Clinical phase molecules
Chen & Mellman (2013) Immunity
T cell killing
Tumor-targeted cytokines
Anti-CEA-IL2v FP
Anti-FAP-IL2v FP
Checkpoint Inhibitors
Anti-PD-L1, Anti-TIGIT,
IDOi, Anti-OX40
11
Atezolizumab strategy: Programs by disease
Focused strategy: Going deep in diseases where
we have strong scientific rationale
15 Roche / Genentech Phase 2 or 3 Studies
LUNG
BLADDER
KIDNEY
2L+ single-arm
Ph2 (x2)
2L+ rand Ph2 and
rand Ph3
1L single-agent
Ph3 (x2)
1L w/chemo,
Avastin Ph3 (x3)
2L+ & 1L cis-inel
single-arm Ph2
1L w/Avastin Ph2
2L+ rand Ph3
1L w/Avastin Ph3
Adjuvant Ph3
BREAST
1L w/Abraxane Ph3
12
Atezolizumab strategy: Summary
Comprehensive, science-driven program
Dx + Monotherapy
Examples:
•  mRCC (Ph II)
•  mUBC (Ph II)
•  2L mNSCLC (Ph II ‘BIRCH’, ‘POPLAR’, ‘FIR’,
Ph III ‘OAK’)
•  Dx+ 1L NSCLC (2x: Sq/NSq, Ph III)
•  Dx+ mUBC (Ph III)
Strategic pillars of our PD-L1
development approach
Dx+
Monotherapy
Chemotherapy
Combinations
Immune Doublets
Examples:
•  Loc adv/ metastatic solid tumors with
ipilimumab or Interferon alfa-2b (Ph Ib)
•  Combination with anti-CSF-1R (Ph Ib)
•  Combination with anti-CD40 (Ph Ib)
•  Combination with Incyte’s IDOi (Ph Ib)
•  Combination with anti-OX40 (Ph Ib)
•  Combination with CEA-IL2v (Ph Ib)
•  Combination with Celldex anti-CD27
Immune
Doublets
Targeted
Therapy
Combinations
Combination with
Chemotherapy
Examples:
•  1L NSCLC Squamous and NonSquamous with platinum doublets (x3,
Ph III)
•  1L TNBC with Abraxane (Ph III)
Combination with
Targeted Therapy
Examples:
•  mRCC with Avastin (Ph II)
•  mRCC with Avastin (Ph III)
•  EGFR+ NSCLC w/Tarceva (Ph Ib)
•  ALK+ NSCLC w/alectinib (Ph Ib)
•  Solid tumors with Avastin (Ph Ib)
•  Solid tumors w/cobimetinib (Ph Ib)
•  Lymphoma with Gazyva (Ph Ib)
13
Strategy in action - Immunotherapy
Perspectives in key franchises
Summary
14
Hematology strategy on track
Strong pipeline and programs across key indications
Compound
Combination
Indication
Gazyva
+chemo GREEN
CLL
Gazyva
+chemo GOYA
aNHL
Gazyva
+chemo GADOLIN
iNHL
Gazyva
+chemo GALLIUM
1L FL
Gazyva
+aPD-L1
R/R FL
Gazyva
+aPD-L1
aNHL
venetoclax*
+Rituxan MURANO
R/R CLL
venetoclax
+Gazyva CLL14
CLL
* venetoclax in collaboration with AbbVie
BR = bendamustine+Rituxan
venetoclax (Bcl2 inhibitor); polatuzumab vedotin (CD79b ADC)
P2
P3
R/R CLL
17p
venetoclax
Biosimilars delayed to 2017
P1
venetoclax
+Rituxan/+BR
R/R FL
venetoclax
+Rituxan/Gazyva+chemo
1L aNHL
venetoclax
+BR
R/R NHL
venetoclax
+bortezomib+chemo
R/R MM
venetoclax
+chemo
AML
polatuzumab +Rituxan/Gazyva ROMULUS NHL
polatuzumab +Gazyva/BR
R/R FL
polatuzumab +Gazyva/BR
aNHL
Data presented at ASCO
15
HER2 franchise outlook
Complete portfolio of innovative drugs
Biosimilar
launch (EU)
2nd line
mBC
Xeloda + lapatinib
1st line
mBC
Herceptin
+ chemo
Adjuvant
BC
Herceptin +
chemo
Neoadjuvant
BC
Herceptin + chemo
(NOAH)1
2011
Kadcyla (EMILIA)
Herceptin & Perjeta + chemo
(CLEOPATRA)
Kadcyla & Perjeta
(MARIANNE)
Herceptin sc + chemo
(HannaH)
2012
Herceptin & Perjeta
+ chemo (APHINITY)
Herceptin & Perjeta + chemo
(Neosphere, Tryphaena)
2013
Established standard of care
2014
2015
Kadcyla (KATHERINE)
Kadcyla & Perjeta
(KAITLIN)
Kadcyla & Perjeta
(KRISTINE)
2016
New standard of care
2017
2018
2019
2020
Trials in progress
16
HER2 franchise: Perjeta commercial uptake strong
CHFm
Perjeta
US Neoadj
Approval
Perjeta
Cleopatra
OS Readout
NEOSPHERE
OS Data at this
ASCO
2,200
Kadcyla
Perjeta
Herceptin
2,000
1,800
1,600
0
Q3
Q4
2013
Q1
Q2
Q3
Q4
2014
Q1
2015
Next Milestone: eBC readout (APHINITY) in 2016
17
Avastin franchise
Continues to be the most studied molecule at ASCO
Continued growth fueled by recent
launches in cervical, ovarian cancers
CHFm
7,000
6,000
5,000
4,000
3,000
2,000
1,000
0
2011
2012
2013
2014
Avastin Mesothelioma data presented at ASCO 2015
Sales at 2014 average exchange rates
18
Strategy in action - Immunotherapy
Perspectives in key franchises
Summary
19
ASCO 2015: Roche highlights
•  More than 275 abstracts, Avastin most quoted drug
•  Potential improvements on current standard of care:
–  Cobimetinib / Zelboraf in 1L skin cancer (targeted therapies)
–  Alectinib in 2L ALK mutated lung cancer
•  New treatment options:
–  Atezolizumab in 2L lung cancer (vs chemo)
–  Atezolizumab combinations with chemo in lung and TNBC
–  Avastin in mesothelioma
–  Gazyva in R/R indolent NHL
•  Several filings resulting from data presented at the meeting
Collaborations: Zelboraf with Plexxikon; Cobimetinib iwith Exelixis; Alectinib with Chugai, Gazyva with Biogen Idec; Kadcyla with ImmunoGen
20
Roche’s oncology strategy has remained constant
Today’s presentations highlight the results of our
strategy across key programs
Follow the science
Personalized
healthcare
•  Bringing our best minds
together
•  Science driven programs
•  Differentiation via
biomarkers
•  Experts integrated in R&D
Strong portfolio &
development
•  Comprehensive pipeline
and active partnering
•  Smart trial design
Higher certainty of patient benefit: Better for patients and for reimbursement
21
Working along the cancer immunity-cycle:
Strategies and new agents I
Ira Mellman
Vice President, Cancer Immunology, Genentech
22
The renaissance of immunotherapy is a
revolution for cancer patients
23
Early data suggests that anti-PD-L1/PD-1 is
active across a wide range of tumor types
Lung cancer
Renal cancer
Bladder cancer
Breast cancer
Head & Neck
cancer
Broad activity but only subset of patients benefit
Melanoma
ORR ~10-30%
Hodgkin’s lymphoma
Most patients will need biomarker selection and combination therapy:
Strategy now supported by clinical results
Using patient data to understand cancer
immunity-cycle
Change in sum of longest diameters
(SLD) from baseline (%)
Atezolizumab Phase 1 data:
Urothelial bladder cancer patients
Complete response
Partial response
Stable disease
Progressive disease
100
80
60
Progressive disease (PD)
Why do many patients not respond?
•  No pre-existing immunity?
Stable disease (SD)
What combinations will promote PRs & CRs?
•  Insufficient T cell immunity?
•  Multiple negative regulators?
40
20
0
-20
-40
-60
-80
-100
0
21 42 63 84 105 126 147 168 189 210 231 252 273 294
Time on study (days)
Powles et al., Nature 2014
Monotherapy durable responses (PR/CR)
What are the drivers of single agent response?
How can PRs be enhanced to CRs?
•  Insufficient T cell immunity?
•  Multiple negative regulators?
Combinations of immunotherapeutics
Increasing response rates & benefit by keeping
cancer immunity-cycle turning
Trafficking of
T cells to tumors
4
Priming and activation
(APCs & T cells)
(CTLs)
Avastin®
3
Anti-OX40
Anti-CD27*
Anti-CTLA4*
5
Infiltration of T cells
into tumors
(CTLs, endothelial cells)
T cell bispecifics
ImmTacs
Cancer antigen
presentation
2
6
(dendritic cells/APCs)
1
Release of
cancer cell antigens
(Immune and cancer cells)
Chen & Mellman (2013) Immunity
(CTLs, cancer cells)
Anti-PD-L1/PD-1
Anti-TIGIT
Anti-OX40 (Treg)
vaccines
Anti-CD40
Genentech/Roche programs
*Partnered programs
Recognition of cancer
cells by T cells
7
Anti-CSF-1R
IDO inhibitor
Anti-CEA-IL2v
Killing of cancer cells
(Immune and cancer cells)
26
IDO (indoleamine di-oxygenase)
A suppressor of effector T cells
Adaptive expression of PD-L1
Adaptive expression of IDO
IDO
IFNg-mediated
up-regulation of
tumor IDO
IFNg-mediated
up-regulation of
tumor PD-L1
Kynurenine
Tryptophan
Shp-2
MAPK"
PI3K
pathways"
Shp-2
MAPK"
PI3K
pathways"
T cell
suppression
Pre-clinical data shows promising activity for IDO
inhibition (GDC-0919)
Anti-PD-L1
control
2000
•  T regulatory cells decrease
•  T effector cells show increased
proliferation & CTL function
•  Enhanced anti-tumor efficacy with
anti-PD-L1 and anti–CTLA4
Tumor volume (mm3)
•  Plasma kynurenine decrease
IDOi
Anti-PD-L1
+ IDOi
1500
1000
500
0
0
10
20
30
40
50
Day
Ongoing studies
•  Phase 1a (GDC-0919)
•  Phase 1b combination (GDC-0919 + atezolizumab)
•  Phase 1b combination (INCB024360 + atezolizumab)
GDC-0919 in collaboration with NewLink;
INCB024360 in collaboration with Incyte
28
TIGIT
A second potent negative regulator of T cell responses
Tumor cell
or DC
ITIM
PVR
1
1nm
100nm
•  Human and murine tumor-infiltrating
CD8+ T cells express high levels of TIGIT
•  Antibody co-blockade of TIGIT and PD-L1
elicits tumor rejection in preclinical models
TIGIT
CD226
ITAM
ITIM
3
2
T cell
•  TIGIT selectively limits the effector function of
chronically stimulated CD8+ T cells
•  TIGIT interacts with CD226 in cis and disrupts
CD226 homodimerization
1 Competes with CD226 for PVR
2 Disrupts CD226 activation
3 Directly inhibits T cells in cis
TIGIT=T cell immunoreceptor with Ig and ITIM domains
29
TIGIT
aTIGIT and aPD-L1 combination effective in PD-L1
non-responsive model
Tumor cell
or DC
PVR
1
1nm
100nm
TIGIT
CD226
ITAM
ITIM
3
2
T cell
Median tumor volume (mm3)
ITIM
10000
Control
1000
100
2 Disrupts CD226 activation
Anti-PD-L1
+ TIGIT
Complete remission (CR)
10
1 Competes with CD226 for PVR
Anti-PD-L1
TIGIT
0
10
20
Day
30
40
60
3 Directly inhibits T cells in cis
30
Elevated expression of TIGIT by T cells in human
cancer supports combination with atezolizumab
Lung squamous cell carcinoma
Tumor
Normal
1.00
CD3ε
1000
200
50
ρ=0.82
2
10
50
TIGIT/CD3ε ratio
5000
0.05
0.2
0.1
0.5
500
TIGIT
Johnson et al (2014) Cancer Cell
31
OX40 exhibits a dual mechanism of action
Promote antigen dependent effector T cell activation
and T regulatory cell inhibition
Primary Tumor Challenge (EMT6)
OX40 engagement on
Treg cells
Antigen
Presenting Cell
Treg
OX40L
TCR
OX40
OX40
IFN-γ
Tumor volume (mm3)
3000
OX40L
OX40L
Control
Anti-mouse OX40
2000
1000
0
0
Effector T
cell
Adapted from Nature Rev Immunol. 4:420 (2004)
10
20
30
40
50
Day
32
Increase in T effector cells by aOX40 may create
need to combine with aPD-L1
anti-OX40
OX40L
OX40
IFN-γ"
T cell
IFNγ
PD-L1
increase
PD-1
anti-PD-L1
(or IDOi)
CD8
T cell
receptor
HLA
PD-L1
Tumor
cell
!!!!!!!!!!!!!!!!!!!!!!!anti&OX40!
anti-OX40
control
!!!!!!!!!!!!!!!!!control!
2000
1500
1000
500
0
0
10
20
30
40
50
anti%PD%L1*
anti-PD-L1
control
control*
2500
2000
1500
1000
500
0
0
Day
day
20
40
Day
day
60
3) 3)
Tumor
(mm
TumorVolume
volume (mm
2500
3) 3)
Tumor
(mm
TumorVolume
volume (mm
3
Tumor
(mm
3) )
TumorVolume
volume (mm
Promising pre-clinical efficacy of an
aOX40/aPD-L1 combination model
2500
anti-OX40 + anti-PD-L1
anti%OX40*+*anti%PD%L1*
2000
1500
1000
500
0
0
Jeong Kim et al. AACR 2015
Ongoing studies
•  Phase 1a (MOXR0916)
•  Phase 1b combination (MOXR0916 + atezolizumab)
•  Planned (MOXR0916 + GDC-0919)
MOXR0916 = anti-OX40
20
40
Day
day
60
Combinations with chemotherapy
Induced inflammation & antigen release may enhance
atezolizumab efficacy
Trafficking of
4
T cells to tumors
(CTLs)
Priming and activation
(APCs & T cells)
3
5
Cancer antigen
presentation
2
6
(dendritic cells/APCs)
Infiltration of T cells
into tumors
(CTLs, endothelial cells)
Recognition of cancer
cells by T cells
(CTLs, cancer cells)
Chemotherapy
Vaccines
1
Release of
cancer cell antigens
(Immune and cancer cells)
Chen & Mellman (2013) Immunity
7
Killing of cancer cells
(Immune and cancer cells)
35
Combinations may extend the benefit of aPD-L1
Preclinical synergy with selected chemo and targeted
therapies
Oxaliplatin chemotherapy
MEKi targeted therapy
CT26 08-1033O
Oxaliplatin
3000
Control
2500
aPD-L1
Tumor Volume (mm3)
2000
1500
1000
Control
2000
Tumor volume, mm3
3
Tumor
volume, mm3
Tumor Volume, mm
2500
CT26 (Krasmut)
MEKi aPD-L1
1500
1000
500
500
aPD-L1/ MEKi
combo
0
0
0
6
12
16
22
29
36
0
Day
Dosing
10
20
30
40
Day
Dosing
control
GDC-0973 (7.5mg/kg)
aPDL1
GDC-0973 (3mg/kg) + aPDL1
50
Combinations with chemotherapy appear to extend
the benefit of atezolizumab in NSCLC patients
Change in Sum of Largest Diameters from Baseline (%)
Atezolizumab +
carboplatin/paclitaxel
100
90
80
70
60
50
40
30
20
10
0
-10
-20
-30
-40
-50
-60
-70
-80
-90
-100
Atezolizumab + carboplatin/
pemetrexed
CR/PR (n=3)
SD (n=2)
PD (n=0)
0
42
84
126
168
210
252
294
Atezolizumab +
carboplatin/nab-paclitaxel
CR/PR (n=9)
SD (n=1)
PD (n=1)
0
42
84
126
168
210
252
294
CR/PR (n=8)
SD (n=4)
PD (n=1)
0
42
84
126
168
210
252
Time on study (days)
Chemotherapy can promote Th1-type inflammation in tumors
Liu et al. ASCO 2015
294
336
Strategic vision: Lead by developing best-in-class
combination therapies
4
Priming & activation
anti-CEA-IL2v
anti-FAP-IL2v
anti-OX40
anti-CTLA4*
anti-CD27*
anti-cytokine
T cell trafficking
Combinations with immunotherapies
aCD40
✔
Vaccines, Oncolytic Viruses
✔✔
aCTLA4
✔
aOX40
✔
aCD27
✔
aCEA/FAP-IL2v
✔
T Cell Bispecifics
✔
ImmTACs
Planned
IDOi
✔✔
aCSF1R
✔
aTIGIT
Planned
Cytokines, anti-cytokines
✔
Combinations with other agents
aVEGF
✔
EGFRi
✔✔
ALKi
Planned
BRAFi
✔
MEKi
✔
BTKi
✔
aCD20
✔
aHER2
✔
Chemo
✔✔
HDAC
✔
A2V
✔
3
5
Immunosuppression
2
Antigen presentation
anti-CD40
IMA942 vaccine*
oncolytic viruses*
neo-epitope vaccine
1
Antigen release
Pro-inflammatory
EGFRi
ALKi
BRAFi
MEKi
Chemo
BTKi
HDAC
T cell infiltration
anti-VEGF
anti-Ang2/VEGF
6
Cancer cell recognition
anti-CEA/CD3 TCB
anti-CEA-IL2v FP
anti-HER2/CD3
7
anti-CD20/CD3
anti-FAP-IL2v
FP
Cancer cell killing
ImmTAC*
atezolizumab
anti-OX40
anti-CSF-1R
IDOi*
anti-TIGIT
IDO/TDO
TDO
✔ Partnered external combo
✔ Internal combo
*
Clinical development
Preclinical development
Partnered projects
Working along the cancer immunity-cycle:
Strategies and new agents II
William Pao, M.D., Ph.D.
Global Head Oncology Discovery and Translational Area, Roche pRED
39
Cancer immunotherapy at pRED
Distinct MoAs with potential for combinations
Macrophage
aCSF-1R
aCEA-IL2v FP
aCEA/CD3 TCB
aCD40
First antibody shown to
eliminate tumorassociated macrophages
Designed to amplify
immune response
First anti-tumor T-cell
engager from Roche
to enter clinical trials
Best/first in class
immuno-activating Ab
Proof of biological
activity established in
rare disease (PVNS)
Novel recombinant
immunocytokine with
superior safety, PK, and
tumor targeting
Best-in-class TCB
platform
Activates antigen presenting cells to
jumpstart adaptive
immunity
40
Tumor-associated macrophages
A promising new target in oncology
41
Emactuzumab (aCSF-1R)
Proof of biological activity shown in PVNS*
Clinical benefit
Emactuzumab Phase I data: PVNS
%
Objective response
24/28
86
Complete response
2/28
7
Partial metabolic
response
24/26
92
Clinically
progression-free
27/28
96
Each color represents a patient
Change in sum of longest diameters
from baseline %
Patients
Time from baseline (days)
*PVNS - pigmented villonodular synovitis, emactuzumab = anti-CSF-1R (RG7155)
CSF1R Inhibition with Emactuzumab in Locally Advanced Diffuse-type Tenosynovial Giant Cell Tumors of the soft tissue: a phase 1, dose
escalation and expansion study, Cassier P. et al. (2015), Lancet Oncology, in press
42
Oral presentation on Mon 1 Jun; abstract #3005
Emactuzumab: Elimination of tumor-associated
macrophages in solid tumors – phase I data
Macrophages suppress T cell
proliferation in vitro
% Change from baseline macrophage
coverage (IHC)
Depletion of CSF-1 dependent
macrophages in solid tumors
RG7155 plus Paclitaxel
RG7155 monotherapy
•  Proof of biological activity in wide range of tumor types
•  Potentially synergistic with atezolizumab
•  Phase I combination study: dose escalation with atezolizumab ongoing
•  No severe immune-related AE reported to date
•  Multiple extensions to start Q3 2015
atezolizumab = anti-PD-L1 (MPDL3280A)
Gomez-Roca, CA: Phase I study of RG7155, a novel anti-CSF1R antibody, in patients with advanced/metastatic solid tumors, 2015 ASCO
43
Tumor targeted IL2v cytokine fusion proteins
A platform to amplify and activate immune effectors
in tumors
Mechanism of Action
•  Binds to target to deliver novel variant of IL-2 to the tumor
•  Recruits killer T and Natural Killer (NK) cells to the tumor
Features
•  Compared to standard IL-2-based therapy, it shows:
−  Superior expansion of immune effectors
−  Less activation of suppressive T-cells
“IL-2v”
−  Better tolerability
−  Better tumor targeting
IL2v – interleukin-2 variant
Novel tumor-targeted, engineered IL-2 variant (IL-2v)-based immunocytokines for immunotherapy of cancer:
Christian Klein et al., ASH, 2013, Blood 122 (21), 2278-2278
44
Oral poster presentation on Saturday 30 May; abstract # 3016*
aCEA-IL2v – phase I data
Tumor targeting and intra-tumoral T cell activation
CEA-IL2v tumor accumulation
shown in 4/4 patients with CEA+ tumors
C1D2
C1D5
CEA +ve
NSCLC
CRC
Baseline
FDG-PET/CT
89Zr-CEA-IL2v-PET/CT
CD8:CD4 Ratio
CD8 density (cells/mm2)
Increased CD8 densities and CD8:CD4 ratio confirm intra-tumoral immune activation at doses ≥ 20 mg
baseline (BL)
4d p.1. CEA-IL2v dose
baseline (BL)
4d p.1. CEA-IL2v dose
aCEA-IL2v is an ideal combination partner for atezolizumab: Phase I combination starting
CEA - carcinoembryonic antigen; *Quantitative data from 2 CEA-ve patients pending, no tumor uptake on visual assessment
*Schellens, Jan H. M. et al., CEA-­targeted engineered IL2: Clinical confirmation of tumor targeting and evidence of intra­tumoral immune
activation; 2015 ASCO
45
aFAP-IL2v: A complementary approach targeting
tumor stroma
FAP staining
Parameter
CEA-IL2v
FAP-IL2v
Target cell
Cancer cells
(CEA)
Tumor stromal cells
(FAP)
Target
expression
Heterogeneous
Homogeneous
Target density
Medium
High
Indications
Mainly GI
indications:
GC, PaC, CRC
Very broadly expressed:
High expression in e.g.
H&N, sq NSCLC, BC
•  FAP-IL2v phase I start: planned in 2015
FAP – fibroblast-activation protein
46
T cell engaging therapies in development
CAR T cells versus T cell bispecific antibodies
Engineered T cells (CARS)
ü
High interest, outstanding clinical efficacy
with CD19 CARs in hematology
û
Activity associated with high toxicity
û
Challenging manufacturing and
regulatory processes
T cell engaging bispecific antibodies
BITE
“1:1”
format
üüü
ü
üüü
Long half-life
û
üüü
üüü
Differentiation of
high vs low
antigen
expressing cells
ü
ü
üüü
Potency
“2:1”
format
47
CEA-TCB: First “ 2:1” T cell bispecific in the clinic
Targets CEA-expressing tumors to induce localized
T cell-mediated killing
Mechanism of Action
High avidity
binding to tumor
antigen
•  Binds T cells and tumor cells simultaneously
•  Results in T cell activation/proliferation and killing of tumor cells
•  Does not require MHC:peptide complex presentation by tumor
cells
•  T cell engagement independent of specificity and activation
status
CD3e T cell
engagement
Low affinity
Inert, engineered
Fc region
Features
•  Entry into human – study started in Dec 2014
•  Combination with IL2v & PD-L1 planned early in clinical
development
•  Broad, early pipeline of several TCBs addressing solid &
hematological malignancies
48
CEA-TCB: Novel mode of action
From poorly inflamed to highly inflamed tumors
Control
CEA-CD3
Intravital two-photon imaging 24 h after CEA-CD3 therapy. Red: Tumor cells (LS174T RFP). Green: Human T cells
•  Recruits new T cells and/or expands pre-existing ones
•  Induces T cell re-localization from tumor periphery to tumor bed
•  Activates intra-tumor T cells to become differentiated towards effector memory phenotype
Bacac M. et al., submitted
49
Roche cancer immunotherapy pipeline
Abundant opportunities for combinations
T cell Trafficking
pRED
gRED
Priming & activation
Eliminate M2
Macrophages
Anti-CSF-1R
Both
T cell infiltration
Costimulatory
antibodies
Anti-OX40 (agonist)
T cell bispecifics
Anti-CEA/CD3 TCB
Antigen presentation
APC stimulators
Anti-CD40 (agonist)
Cancer T cell recognition
Antigen release
Clinical phase molecules
Chen & Mellman (2013) Immunity
T cell killing
Tumor-targeted cytokines
Anti-CEA-IL2v FP
Anti-FAP-IL2v FP
Checkpoint Inhibitors
Anti-PD-L1, Anti-TIGIT,
IDOi, Anti-OX40
50
Cancer immunotherapy program growing strongly
Compound
Combination
Mono
+Tarceva
+chemo
+Avastin+chemo
Indication
aPD-L1
Mono
Bladder cancer
aPD-L1
Mono
+Avastin
Renal cancer
aPD-L1
+Zelboraf
+Zelboraf+cobimetinib
Melanoma
aPD-L1
Mono
+Avastin
+cobimetinib
+ipilimumab
+IFN alfa-2b
+aCD40
+aOX40
+aCSF-1R
+aCEA-IL2v FP
aPD-L1
+Avastin+FOLFOX
Colorectal cancer
aPD-L1
Mono
+Gazyva
Hematology
aPD-L1
Mono
+chemo
Triple neg. breast cancer
aCSF-1R
Mono
+aCD40
Solid tumors
aCEA-IL2v FP
Mono
Solid tumors
aFAP-IL2v FP
Mono
Solid tumors
aOX40
Mono
Solid tumors
aCEA/CD3 TCB
Mono
Solid tumors
IDO
Mono
Solid tumors
IDO (Incyte)*
+aPD-L1
Solid tumors
Varlilumab (Celldex)*
+aPD-L1
Solid tumors
aPD-L1
Phase 1
Phase 2
Phase 3
Lung cancer
Solid tumors
Study ongoing
51
Study imminent
* External collaboration
ASCO 2015 Roche highlights:
Setting new standards, developing combinations
Sandra Horning, M.D.
Chief Medical Officer and Head Global Product Development
52
Agenda
Setting new standards, developing combinations
Cancer immunotherapy
Roche program update
Atezolizumab in lung cancer
Atezolizumab in bladder cancer
Targeted therapies and combinations
Alectinib in ALK-positive lung cancer
Cobimetinib + Zelboraf in melanoma
Kadcyla, Perjeta, Herceptin in HER2+ breast cancer
Hematology
Gazyva in NHL
53
Roche cancer immunotherapy program
Phase I
üü
ü
aPD-L1
Solid tumors
aPD-L1+chemo
Solid tumors
aPD-L1+lenalidomide**
MM
aPD-L1+Tarceva
NSCLC
aPD-L1+Zelboraf
Melanoma
aPD-L1+cobimetinib
Solid tumors
aPD-L1+Avastin
Solid tumors
aPD-L1 + Gazyva
R/R FL / aNHL
aPD-L1+Avastin+chemo
Solid tumors
aPD-L1+Zelboraf+cobi**
Melanoma
aCSF-1R
Solid tumors
aCEA-IL2v FP
Solid tumors
aOX40
Solid tumors
aCEA/CD3 TCB
Solid tumors
ü
ü
Status as at May 31, 2015
Phase II
IDO
Solid tumors
aPD-L1+ipilimumab
Solid tumors
aPD-L1+IFN-alfa
Solid tumors
aPD-L1+aCD40
Solid tumors
aPD-L1+aOX40**
Solid tumors
aPD-L1+aCSF-1R**
Solid tumors
aPD-L1
NSCLC (Dx+)
aPD-L1
2/3L NSCLC
aPD-L1+Avastin
1L Renal
aPD-L1
1/2L Bladder
Phase III
ü
ü
aPD-L1+aCEA-IL2v FP**
Solid tumors
aPD-L1**
tba
aPD-L1**
tba
aPD-L1**
tba
aPD-L1**
tba
aPD-L1**
tba
NME**
tba
aPD-L1 trials
NMEs monotherapy
aPD-L1
2/3L NSCLC
aPD-L1**
2/3L Bladder
aPD-L1+Avastin+chemo**
1L non sq NSCLC
aPD-L1+chemo**
1L non sq NSCLC
aPD-L1+chemo**
1L sq NSCLC
aPD-L1**
1L non sq NSCLC (Dx+)
aPD-L1**
1L sq NSCLC (Dx+)
aPD-L1+chemo**
1L TNBC
aPD-L1+Avastin**
1L RCC
aPD-L1**
Adjuvant MIBC (Dx+)
aPD-L1**
tba
Immune doublets
2015 readout expected
**
Study start in 2015
ü
Data at ASCO 2015
54
Biomarker development for atezolizumab
PD-L1 expression on different cell types
PD-L1 expression on immune cells (IC) and tumor cells (TC)
IC Score PD-L1 staining TC Score PD-L1 staining IC3 IC ≥ 10% TC3 TC ≥ 50% IC2 IC ≥ 5% and < 10% TC2 TC ≥ 5% and < 50% IC1 IC ≥ 1% and < 5% TC1 TC ≥ 1% and < 5% IC0 IC < 1% TC0 TC < 1% Assessed by immunohistochemistry using SP142, a sensitive and
specific proprietary monoclonal antibody assay against PD-L1
Immune cell staining
Gettinger S. et al, ASCO 2015
Tumor cell staining
55
Atezolizumab clinical program in 2/3L NSCLC
Breakthrough designation in PD-L1+ patients
Primary end-point:
FIR
Phase II
PD-L1-selected
atezolizumab
mNSCLC n=138
1200 mg IV Q3 weeks
All comers
atezolizumab
atezolizumab
1200 mg
mg IV
IV Q3
Q3 weeks
weeks
1200
POPLAR Phase II
2/3L mNSCLC n=287
PD-L1 stratified
docetaxel
ORR
Data at ASCO
OS
Interim data at ASCO
Final data in H2 2015
75 mg/m2 IV Q3 weeks
BIRCH
Phase II
atezolizumab
PD-L1-selected
mNSCLC n=667
OAK
1200 mg IV Q3 weeks
Phase III
ORR
Data in Q3 2015
atezolizumab
All comers
1200 mg IV Q3 weeks
2/3L mNSCLC n=1100
PD-L1 stratified
docetaxel
OS
Data expected 2016
75 mg/m2 IV Q3 weeks
Note: Atezolizumab (Anti-PD-L1) is listed as MPDL3280A in clinicaltrials.gov
mNSCLC = metastatic Non Small-Cell Lung Cancer
56
Atezolizumab in 2/3L NSCLC (POPLAR)
Clinical outcomes correlate with PD-L1 expression
Overall survival
ITT
16%*
TC2/3 or IC2/3
37%*
TC1/2/3 or IC1/2/3
67%*
TC0 and IC0
32%*
0
0.98
0.77
n=287
TC3 or IC3
0.57
0.46
0.70
0.56
0.87
0.63
1.12
1.17
1.87
1.93
0.5
1
HRa
In favor of atezolizumab
= unstratified HR; * = eligible patient population
Spira A. et al, ASCO 2015
a
Progression-free survival
1.5 0
In favor of docetaxel
0.5
1
HRa
In favor of atezolizumab
1.5
In favor of docetaxel
57
Atezolizumab in 2/3L NSCLC (POPLAR)
Overall survival correlates with PD-L1 expression
TC1/2/3 or IC1/2/3
TC0 or IC0
Not Reached
(11.0, NE)
9.1 mo
(7.4, 12.8)
9.7 mo
(6.7, 11.4)
9.7 mo
(8.6, 12.0)
•  Final POPLAR and BIRCH read-outs in H2 15
•  Phase III (OAK) in all comers (stratified by PD-L1 expression): Data in 2016
= Unstratified HR
Spira A. et al, ASCO 2015
a
58
Atezolizumab + chemotherapy in 1L NSCLC (Ph1)
Well tolerated with no unexpected toxicities
1L NSCLC
n=37
4-6 cycles
maintenance
C
atezolizumab 15mg/kg IV q3w+
carboplatin q3w + paclitaxel q3w
atezolizumab
Treat to PD
or
loss of clinical benefit
D
atezolizumab 15mg/kg IV q3w+
carboplatin q3w + pemetrexed q3w
atezolizumab +/pemetrexed
Treat to PD
or
loss of clinical benefit
atezolizumab 15mg/kg IV q3w+
carboplatin q3w + nab-paclitaxel q1w
atezolizumab
Treat to PD
or
loss of clinical benefit
E
•  The addition of atezolizumab to standard 1L chemotherapy was well tolerated
throughout the course of treatment, including maintenance therapy, with no
unexpected toxicities
•  No autoimmune renal toxicity or pneumonitis was observed
Liu S. et al, ASCO 2015
59
Atezolizumab + chemotherapy in 1L NSCLC (Ph1)
Efficacy unrelated to PD-L1 expression
Arm C, n=5
atezolizumab +
carboplatin/pemetrexed
atezolizumab +
carboplatin/nab-paclitaxel
60%
(19-92)
75%
(45-93)
62%
(33-83)
100
90
80
70
60
50
40
30
20
10
0
-10
-20
-30
-40
-50
-60
-70
-80
-90
-100
CR/PR (n=3)
SD (n=2)
PD (n=0)
0
42
Arm E, n=13
atezolizumab +
carboplatin/paclitaxel
ORR
(95% CI)
Change in Sum of Largest Diameters from Baseline (%)
Arm D, n=12
84
126
168
210
252
294
CR/PR (n=9)
SD (n=1)
PD (n=1)
0
42
84
126
168
210
252
294
CR/PR (n=8)
SD (n=4)
PD (n=1)
0
42
84
126
168
210
252
294
336
Time on study (days)
ORR across all arms = 67% (48-82)
CR = complete response; PR = partial response; SD = stable disease; PD = progressive disease
Liu S. et al, ASCO 2015
60
Atezolizumab phase III program in 1L NSCLC
Developing combinations, building on predictive
PD-L1 biomarker
Study
Indication
Treatment arms
Primary
completion*
Combination studies – All comers (PD-L1 subgroup analysis)
GO29436
n=1200
GO29537
n=550
GO29437
n=1200
Non-squamous
Atezo + Carbo + Pac
Atezo + Carbo + Pac + Avastin
Carbo + Pac + Avastin
2017 (PFS)
Non-squamous
Atezo + Carbo + Nab-pac
Carbo + Nab-pac
2017 (PFS)
Squamous
Atezo + Carbo + Pac
Atezo + Carbo + Nab-pac
Carbo + Nab-pac
2017 (PFS)
Non-squamous
Atezo
Cis or Carbo + Pem
2017 (PFS)
Squamous
Atezo
Cis or Carbo + Gem
2017 (PFS)
Monotherapy studies – PD-L1 Selected
GO29431
n=400
GO29432
n=400
* Outcome studies are event driven, timelines may change, OS endpoint included for all studies
Atezo=atezolizumab; Carbo=Carboplatin; Pac=Paclitaxel; Nab-pac= Nab-paclitaxel; Cis=Cisplatin; Pem=Pemetrexed; Gem=Gemcitabine
Note: Atezolizumab (Anti-PD-L1) is listed as MPDL3280A in clinicaltrials.gov `
61
Atezolizumab in 2L UBC (Ph1)
ORR correlates with PD-L1 expression
Atezolizumab responses by PD-L1 biomarker status
PD-L1 IHC
n=87
ORR, % (95% CI)
IC3 (n=12)
67% (35, 90)
IC2 (n=34)
44% (27, 62)
IC1 (n=26)
19% (7, 39)
IC0 (n=15)
13% (2, 40)
50% (35, 65)
IC3
IC2
IC1
IC0
17% (7, 32)
•  Response rates correlate with PD-L1 status; 20% CR in IHC2/3
•  Responses observed with visceral metastases: 38% RR in IHC2/3
•  Well tolerated with 5% Gr3/4 immune-related adverse events
UBC = urothelial bladder Cancer; IC = immune cells; ORR = overall response rate
Petrylak D. et al, ASCO 2015
62
Atezolizumab in 2L UBC (Ph1)
Survival correlates with PD-L1 expression
Overall survival
Survivala
N = 92
IC2/3
n = 48
Not Reached
IC0/1
(95% CI, 9.0-NE)
n = 44
OS
Median OS
(range)
1-y survival
(95% CI)
Not reached
8 mo
(1 to 20+ mo)
(1 to 15+ mo)
57%
38%
(41-73)
IC2/3
IC0/1
(19-56)
7.6 mo
(95% CI, 4.7-NE)
• 
PD-L1 is a predictive biomarker for ORR, PFS and OS in 2L bladder cancer
• 
Phase II study (NCT02108652) in 1/2L bladder: Read-out H2 2015
• 
Phase III studies in 2/3L bladder and high risk muscle invasive bladder after cystectomy (adjuvant)
started in 2015
Petrylak D. et al, ASCO 2015
63
Agenda
Setting new standards, developing combinations
Cancer immunotherapy
Roche program update
Atezolizumab in lung cancer
Atezolizumab in bladder cancer
Targeted therapies and combinations
Alectinib in ALK-positive lung cancer
Cobimetinib + Zelboraf in melanoma
Kadcyla, Perjeta, Herceptin in HER2+ breast cancer
Hematology
Gazyva in NHL
64
Alectinib in ALK+ NSCLC after crizotinib failure
Strong efficacy in two phase 2 studies
Results from P2 US and Global Studies
•  Alectinib, next generation, brain penetrant
ALK inhibitor
Overall and CNS Response Rates
80
69%
70
57%
60
50
50%
46%
ORR
CR
40
All
(n=69)
30
20
10
0
20%
13%
All
(n=87)
Measurable
CNS disease
(n=16)
US Study
All
(n=122)
Measurable
CNS disease
(n=16)
•  Two P2 studies (global and US) show high
systemic and CNS response rates in
crizotinib-failed patients
•  Median PFS: 8.9m & 6.3m (global & US
study, respectively)
•  Good tolerability with minimal dose
modifications
•  Final duration of response and PFS results
expected in H2 2015
Global Study
Ou S-H. et al, ASCO 2015; Gandhi L. et al, ASCO 2015
65
Alectinib in ALK+ NSCLC after crizotinib failure
Excellent CNS disease control
Measureable CNS responses in global study
Day 6
Day 87
•  Both global and US studies show high ORR (57-69%) in measureable CNS disease; disease control
rates 80-100%
•  Median duration of response in measureable CNS disease was 10.3 months in the global study
•  Filing planned 2015
•  Global phase III study in 1L ALK+ NSCLC (Alex) on-going: alectinib vs crizotinib
Ou S-H. et al, ASCO 2015; Gandhi L. et al, ASCO 2015
66
Cobimetinib+Zelboraf in Melanoma (coBRIM)
Update confirms benefit in BRAFV600+ patients
Orpha
disea n
desig se
natio
n
Progression free survival
HR = 0.58 (0.46-0.72)
Zelboraf +
Cobimetinib
(n = 247)
Zelboraf +
Placebo
(n = 248)
% ORR
69.6
50.0
% CR
15.8
10.5
Median PFS
12.3
7.2
Median DoR in
months
13.0
9.2
7.2m
12.3m
•  Median PFS of ~1 year in Cobimetinib + Zelboraf arm with longer follow-up (14.2 m)
•  OS data not yet mature
•  Filed in US and EU
ORR = overall response rate; CR = complete response; PR = partial response; DoR = duration of response
Larkin J. et al, ASCO 2015
67
Cobimetinib+Zelboraf in Melanoma (BRIM7)
Phase I update shows strong efficacy
Overall survival
Zelboraf Progressors (N=66)
BRAF inhibitor
−naive
(n=63)
ZelborafProgressors
(n=66)
ORR%
87.3
15.2
CR%
15.9
1.5
mPFS (m)
13.8
2.8
mOS (m)
28.5
8.4
1-yr OS%
82.5
34.7
2-yr OS%
61.1
15.1
Vem prog
Naive
1
5
9
13
17
21
25
29
33
37
Time (months)
•  Two-year survival rate demonstrates benefit of cobimetinib + Zelboraf
•  Projected 28.5 m median OS compares to historical 13.6 m with Zelboraf alone
•  Results show importance of combination vs sequential use of targeted therapy in melanoma
CR = complete response; PR = partial response; SD = stable disease; PD = progressive disease
Pavlick A. et al, ASCO 2015
68
Kadcyla + Perjeta in 1L HER2+ mBC (MARIANNE)
Non-inferior PFS for Kadcyla or Kadcyla + Perjeta
Median
PFS (m)
HR vs
H+T
HR vs
K
Herceptin + docetaxel (H+T)
(8 mg/kg LD then 6 mg/kg + 100 or 75 mg/m2 q3w)
OR
Herceptin + paclitaxel (H+T)
Previously
untreated
HER2+mBC
n=1092
13.7
(4 mg/kg LD then 2 mg/kg + 80 mg/m2 qw)
1:1:1
Kadcyla + placebo (K)
14.1
Kadcyla + Perjeta (K+P)
15.2
(3.6mg/kg + 840 mg LD then 420 mg q3w)
(3.6mg/kg + 840 mg LD then 420 mg q3w)
0.91
p=0.31
0.87
p=0.14
0.91
•  K and K + P achieved non-inferiority but not superiority to H + T
•  Addition of Perjeta to Kadcyla did not improve PFS
•  Kadcyla was better tolerated and maintained health-related quality of life for a longer duration
Ellis P. et al, ASCO 2015
69
Perjeta + Herceptin in HER2+ eBC (NEOSPHERE)
5yr follow-up shows lasting benefit in neoadjuvant setting
neoadjuvant
adjuvant
Herceptin+docetaxel
(H+T)
n=107
Perjeta+Herceptin+docetaxel (P+H
+T)
n=107
Perjeta+Herceptin
n=107
S
U
R
G
E
R
Y
pCR
FEC q3 w x 3*
+ Herceptin
Perjeta+docetaxel
n=96
Perjeta + Herceptin + docetaxel
•  PFS and DFS benefit of neoadjuvant Perjeta +
Herceptin + docetaxel - despite identical
adjuvant therapy
•  No new safety concerns with longer follow-up
PFS,
DFS
P+H+T
(n=107)
H+T
(n=107)
Events PFS
17
19
5-year PFS
86%
81%
HR
0.69 (0.34-1.40)
Events DFS
15 (14.9%)
18 (17.5%)
5-year DFS
84%
81%
HR
0.60 (0.28-1.27)
pCR endpoint
•  Achievement of tpCR (ITT analysis of all
patients) reduced risk of PFS (HR = 0.54)
•  Results support pCR as an earlier indicator of
benefit
•  APHINITY adjuvant study results in 2016
FEC = 5-fluorouracil, Epirubicin, Cyclophosphamide; pCR = pathologic complete response; PFS = progression free survival;
DFS = disease free survival
Gianni L. et al, ASCO 2015
* PT arm also received docetaxel q3w X 4
70
HER2 franchise outlook
Complete portfolio of innovative drugs
Biosimilar
launch (EU)
2nd line
mBC
Xeloda + lapatinib
1st line
mBC
Herceptin
+ chemo
Adjuvant
BC
Herceptin +
chemo
Neoadjuvant
BC
Herceptin + chemo
(NOAH)1
2011
Kadcyla (EMILIA)
Herceptin & Perjeta + chemo
(CLEOPATRA)
Kadcyla & Perjeta
(MARIANNE)
Herceptin sc + chemo
(HannaH)
2012
Herceptin & Perjeta
+ chemo (APHINITY)
Herceptin & Perjeta + chemo
(Neosphere, Tryphaena)
2013
Established standard of care
2014
2015
Kadcyla (KATHERINE)
Kadcyla & Perjeta
(KAITLIN)
Kadcyla & Perjeta
(KRISTINE)
2016
New standard of care
2017
2018
2019
2020
Trials in progress
•  APHINITY, KRISTINE results in 2016
•  NEOSPHERE update ASCO 2015
71
Agenda
Setting new standards, developing combinations
Cancer immunotherapy
Roche program update
Atezolizumab in lung cancer
Atezolizumab in bladder cancer
Targeted therapies and combinations
Alectinib in ALK-positive lung cancer
Cobimetinib + Zelboraf in melanoma
Kadcyla, Perjeta, Herceptin in HER2+ breast cancer
Hematology
Gazyva in NHL
72
Hematology franchise
A broad portfolio of innovative drugs
Compound
Combination
Indication
Gazyva
+chemo GREEN
CLL
Gazyva
+chemo GOYA
aNHL
Gazyva
+chemo GADOLIN
iNHL
Gazyva
+chemo GALLIUM
1L FL
Gazyva
+aPD-L1
R/R FL
Gazyva
+aPD-L1
aNHL
venetoclax*
+Rituxan MURANO
R/R CLL
venetoclax
+Gazyva CLL14
CLL
* venetoclax in collaboration with AbbVie
BR = bendamustine+Rituxan
venetoclax (Bcl2 inhibitor); polatuzumab vedotin (CD79b ADC)
P1
P2
P3
R/R CLL
17p
venetoclax
Biosimilars delayed to 2017
Brea
desig kthrough
natio
n for therapy
vene
in R/R
CLL 1 toclax
7p
venetoclax
+Rituxan/+BR
R/R FL
venetoclax
+Rituxan/Gazyva+chemo
1L aNHL
venetoclax
+BR
R/R NHL
venetoclax
+bortezomib+chemo
R/R MM
venetoclax
+chemo
AML
polatuzumab +Rituxan/Gazyva ROMULUS NHL
polatuzumab +Gazyva/BR
R/R FL
polatuzumab +Gazyva/BR
aNHL
Data presented at ASCO
73
Second positive readout for Gazyva
GADOLIN in Rituxan-refractory iNHL
Primary end-point:
CLL11: Ph III Chronic Lymphocytic Leukemia (CLL)
Gazyva + chlorambucil
1L CLL
n=781
PFS
Approved in Q4 2013
Rituxan + chlorambucil
chlorambucil
GADOLIN: Ph III Recurrent Indolent NHL (iNHL)
Induction
Rituxan-refractory iNHL
n=411
Gazyva + bendamustine
Maintenance
CR, PR,
SD
bendamustine
Gazyva
q2mo x 2 years
PFS
Data at ASCO 2015
GOYA: Ph III 1L Diffuse Large B-cell Lymphoma (DLBCL)
Front-line DLBCL
(aggressive NHL)
n=1418
Gazyva + CHOP
GALLIUM: Ph III 1L Indolent NHL (iNHL)
1L iNHL
n=1401
PFS
Trial to continue until 2016
Rituxan + CHOP
Induction
Gazyva + CHOP or
Gazyva + CVP or
Gazyva + bendamustine
Rituxan + CHOP or
Rituxan or
Rituxan + bendamustine
Maintenance
Gazyva
q2mo x 2 years
CR, PR
Rituxan
q2mo x 2 years
PFS
Enrollment complete Q1 2014
Data expected 2017
In collaboration with Biogen Idec
CHOP=Cyclophosphamide, Doxorubicin, Vincristine and Prednisone; CVP=Cyclophosphamide, Vincristine and Prednisolone
74
Gazyva in Rituxan-refractory iNHL (GADOLIN)
Significant PFS improvement achieved
IFR-assessed PFS
OS
HRa = 0.55 (0.40-0.74)
not reached
14.9m
•  PFS significantly prolonged by independent (HR=0.55) and investigator (HR 0.52) assessment
•  Median PFS was ~ doubled by investigator assessment (29 versus 14 months)
•  OS data immature
•  Filing planned in 2015
= Stratified HR
Sehn L. et al, ASCO 2015
a
75
Setting new standards, developing combinations
Driven by the breadth of our portfolio
CEA-IL2v
chemo
anti-OX40
CSF-1R
Launched
portfolio
Immunotherapy
portfolio
CEA-CD3
atezolizumab
IDO/TDO
anti-CD40
alectinib
venetoclax
cobimetinib
Targeted combinations approved
Chemotherapy combinations approved
Roche combinations in trials
Chemotherapy combinations in trials
NMEs to be filed soon
76
Growing importance of molecular information in
cancer immunotherapy
Garret Hampton, Ph.D.
Vice President, Oncology Biomarker Development and Companion Diagnostics
77
Personalized healthcare
A cornerstone of our strategy
Diagnostics
Molecule
Pharma
Dx
Gazyva
CLL
Alectinib
ALK+ NSCLC
Atezolizumab
PD-L1+ NSCLC
Atezolizumab
4 of 5 BTD
were enabled by
having a Dx
that identified
patients most
likely to benefit
PD-L1+ UBC
Right patient – right drug
Venetoclax
17p- CLL
> 60% of current drugs in our development portfolio have a companion diagnostic
78
Oncology is evolving
More complex science with better patient outcomes
Disease heterogeneity in lung cancer
PIK3CA
HER2
Overall survival in EGFR mut+ lung cancer
AKT1
BRAF
MAP2K1
EGFR
KIF5B-RET
ROS1 Fusions
NRAS
PD-L1
expression
Unknown
KRAS
MET
Amplification
PFS probability
ALK Fusions
1.0
Erlotinib (n = 86)
0.8
Chemotherapy (n = 87)
HR 0.37 (95% CI 0.25–0.54);
0.6
log-rank p < 0.0001
0.4
0.2
0
MET splice site
0
3
9
12 15 18 21 24 27 30 33 36
Time (months)
Number
at risk
Erlotinib 86 63
Chemo 87 49
Rosell et al. (2012) The Lancet
6
54
20
32
8
21
5
17
4
9
3
7
1
4
0
2
0
2
0
0
0
0
0
79
Biomarker discovery and development at
Genentech/Roche
Understand disease
biology
• 
• 
• 
• 
• 
Biomarker prevalence
Prognostic significance
Subsets of unmet need
Pipeline opportunities
Dx hypotheses
Identify
biomarkers
• 
• 
• 
• 
Pharmacodynamic
Predictive
Response surrogate
Resistance (innate/acquired)
Develop new
technologies
•  Multiplex assays
•  Real-time Dx
•  Monitoring
Disease
Technology
Therapeutics
Define diagnostic
endpoints
•  Companion Dx endpoints
•  Response surrogates
80
Predictive markers – tissue staining for PD-L1
Cancer and/or immune cell expression of PD-L1
can correlate with clinical benefit
n
% of PD-L1–positive
(immune cells)
% of PD-L1–positive
(tumor cells)
NSCLC
184
26
24
RCC
88
25
10
Melanoma
59
36
5
HNSCC
101
28
19
Gastric
141
18
5
CRC
77
35
1
Pancreatic
83
12
4
Bladder cancer
205
27
11
Tumor type
Herbst et al (2014) Nature
PD-L1+ (immune cells)
PD-L1+ (tumor cells)
81
Patient selection enriches for benefit
PD-L1-selected lung (TC & IC) and bladder cancer (IC)
PD-L1 TC staining
Score
PD-L1 IC staining
Score
TC ≥ 50%
TC ≥ 5% and < 50%
TC ≥ 1% and < 5%
TC < 1%
TC3
TC2
TC1
TC0
IC ≥ 10%
IC ≥ 5% and < 10%
IC ≥ %1 and < 5%
IC < 1%
IC3
IC2
IC1
IC0
Lung cancer: Survival hazard ratio*
100
0.46
TC3 or IC3
80
Overall survival
0.56
TC2/3 or IC2/3
0.63
TC1/2/3 or IC1/2/3
1.12
TC0 and IC0
0.77
ITT (N = 287)
0.1
0.2
1
1
Hazard Ratioa
2
In favor of atezolizumab In favor of docetaxel
* Monotherapy data
Bladder cancer: Overall survival*
Median OS Not Reached
(95% CI, 9.0-NE)
60
40
20
0
IC2/3
IC0/1
+ Censored
Median OS 7.6 mo
(95% CI, 4.7-NE)
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
Time (months)
82
Using patient data to understand cancer
immunity-cycle
Change in sum of longest diameters
(SLD) from baseline (%)
Atezolizumab Phase 1 data:
Urothelial bladder cancer patients
Complete response
Partial response
Stable disease
Progressive disease
100
80
60
Progressive disease (PD)
Why do many patients not respond?
•  No pre-existing immunity?
Stable disease (SD)
What combinations will promote PRs & CRs?
•  Insufficient T cell immunity?
•  Multiple negative regulators?
40
20
0
-20
-40
-60
-80
-100
0
21 42 63 84 105 126 147 168 189 210 231 252 273 294
Time on study (days)
Powles et al., Nature 2014
Monotherapy durable responses (PR/CR)
What are the drivers of single agent response?
How can PRs be enhanced to CRs?
•  Insufficient T cell immunity?
•  Multiple negative regulators?
83
Biomarker discovery and development
Integral part of the cancer immunotherapy
learning organization
Research
(gRED, pRED, Chugai & external)
Development
(early and late)
Cancer Immunotherapy
Committee
Partnering
(Genentech and Roche)
Clinical Insights &
Biomarkers
(internal and external)
84
Combination with Avastin
Increasing response rates by keeping cancer
immunity-cycle turning
On-treatment biopsies show absence of T cells in
tumor bed before and after treatment with PD-L1
Pre-treatment
On-treatment (week 6)
CD8
Patient 1
CD8
Lack of T cell infiltration may limit response to
checkpoint inhibitors
Anti-VEGF
CD8
Patient 2
CD8
Herbst et al. (2014) Nature
Chen & Mellman (2013) Immunity
85
Combination with Avastin
Increased T cell infiltration and clinical activity
CD31
(vasculature)
CD8
(T cells)
Pre-treatment
Avastin
Avastin + aPD-L1
Combination regimen benefits most patients irrespective
of PD-L1 status
100
90
80
70
60
50
40
30
20
10
0
-10
-20
-30
-40
-50
-60
-70
-80
-90
-100
PR
SD
PD
Change in sum of largest diameters
from baseline (%)
On-treatment biopsies show increased infiltrate and
reduction in tumor vasculature
IHC 3
IHC 0
IHC 0
IHC 1
IHC 1
IHC 0
IHC 0
IHC 1
IHC 0
Discontin.
0
42
84
126
168
210 252
294 336
378
Time on study (days)
Phase 2 in RCC ongoing (n= 300); Phase 3 initiated
Sznol et al. ASCO GU 2015
86
Clinical data guides combinations
Identifying potentially synergistic combinations
Myeloid signature associated with lack of response
to atezolizumab in bladder cancer
, Th2 and Myeloid signature by response group. Red: PD; Black: SD; Blue: CR/PR.
Immune marker expression
(illustrative)
Th2 signature
N=15
N=23
●
●
●●
●
●
●
●
●
●
●
●●
●
●
●
●
−1.6
−0.67
1
TNBC
Adeno
NSCLC
Bladder
RCC
SCC
NSCLC
Melanoma
−2
PD
0.22
0
1.1
1.6
SD
2
Th2 Signature
●
Treg
● Signature
●
●
●
●
●
Expression
−16
●
●●
● Signature
Teff
0
●
●
●●
●
●
●●
●
●
●●
●
CR/PR
●
IFNg Signature
●
●
IB●●
Signature
●
●
B cell
● Signature
Th17 Signature
●
●
CR/PR
−2
−14
Expression
●
NK cell Signature
●●
−18
●
APC Signature
●
●
●
Myeloid Signature
●
4
●
cbind(1:nc)
−12
●
●●
●
2
N=23
N=20
N=15
●
CRC
●
●
●
●
N=15
HR+BC
●● ●
●
N=20
Myeloid signature:
IL1B,
IL8,signature
CCL2
Myeloid
●
Indication
−10
●
HER2+BC
e
Complementary mechanisms of action to enhance
benefit from immunotherapies
●
●
●●
●
●
●●
●
●●
● ●●
●●
●
●
●
●
●
●
P=0.01
PD
●●
●
●
●●
●
●
●
●
●●
●
●
●
●
●
●
●●
●
●
SD
Anti-CSF-1R
●
●
●
●
●
●
●●
●
●●
●
●
Anti-PD-L1
CR/PR
2
-1 -2
Progressive disease
Tumor associated macrophages
30
Phase I combination study of anti-CSF-1R and atezolizumab ongoing
87
Roche and FMI will innovate together
Exclusive development of immunotherapy panel
Tumor profiling with RNA sequencing
Mutation-load and neo-antigen discovery
oxplot of Te↵, Th2 and Myeloid signature by response group. Red: PD; Black: SD; Blue: CR/PR.
Myeloid signature:
IL1B,Myeloid
IL8,signature
CCL2 .
Immune marker expression
(illustrative)Th2 signature
Teff signature
●●●
●
●
●
−2
2
−1.6
−0.67
1
0.22
0
1.1
●●
●
●●
1.6
CRC
TNBC
HER2+BC
PD
APC Signature
●
● IFNg Signature
●
●
● IB Signature●●
●
●
●
●●
●
● B cell Signature
●
●
●
● Teff Signature
●
●
Th2 Signature
●
●●
Treg Signature
●
●
● Th17 Signature
●
●
●
●
●
SD
CR/PR
2
●
NK cell Signature
●
Bladder
CR/PR
RCC
●
SD
SCC
NSCLC
●
Myeloid Signature
4
●
●
Melanoma
D
●
●
●
●●
●
●
●
●
●
●
●
●●
●
●
●
●
Expression
●
●
●
−18
●
●
●
cbind(1:nc)
0
●
N=23
−2
●
●
●●
●
Expression
●
●●
●
N=15
N=20
N=15
●
●
HR+BC
●
●
●●
●
●
●●
●
●
● ●●
●
N=20
●
●● ●
●
●
●
●
●●
N=23
−10
●
N=15
●
Indication
−12
●
−14
N=20
−16
●
●
Adeno
NSCLC
23
●
●
●●
●
●
●●
●
●●
● ●●
●●
●
●
●
●
●
●
P=0.01
PD
●●
●
●
●●
●
●
●
●
●●
●
●
●
●
●
●
●
●●
●
●●
●
●
●
●
●
●
●
●●
●
●
SD
CR/PR
2
-1 -2
Progressive disease
Predictive biomarkers
& new combinations
30
Schumacher and Schreiber (2015) Science
Yadav et al (2014) Nature
Identifying immunogenic mutations by
structure-guided algorithms
88
Our vision for cancer immunotherapy
A personalized approach to treatment
*Evaluate tumor:
Is the tumor inflamed?
Y
N
Inflamed
Non-Inflamed
1
2
3
4
1
2
3
4
Strong PD-L1
Weak PD-L1
No PD-L1
No identifiable
immune targets
Are T cells at
tumor periphery?
MHC loss?
No T cells?
No identifiable
immune targets
Tx with
aPD-L1/PD-1
Are suppressive
myeloid cells present?
IDO/kyneurinin
expressed?
Avastin+
Atezolizumab
Tumor antigen
expression?
Antigen
experienced?
Chemo+
Atezolizumab
Anti-CSF-1R+
Atezolizumab
IDOi+
Atezolizumab
T cell
BiSpecific+
Atezolizumab
Anti-OX40+
Atezolizumab
*Hypothetical algorithm
Chemo+
Atezolizumab
89
Patient journey tomorrow
Bridging comprehensive testing and drug development
Molecular
monitoring
Patient
care
R&D
insights
Clinical
decision
Comprehensive
testing
90
Summary and Q&A
Karl Mahler
Head of Investor Relations, Roche
91
Summary
Skin cancer
cobimetinib + Zelboraf: Ph III (coBRIM) in 1L BRAF+ mM; PFS & biomarker update
Filed in 2014
cobimetinib + Zelboraf: Ph Ib (BRIM7) in BRAF+ mM; OS update
Lung cancer
alectinib: two Ph II in 2L ALK+ NSCLC
Filing in 2015
Atezolizumab (anti-PDL1): POPLAR, COMBO, FIR
Discuss filing 2015
Avastin: mesothelioma
Discuss filing 2015
Bladder cancer
Atezolizumab (anti-PDL1): Ph I update in bladder
Discuss filing 2015
Breast cancer
Kadcyla: Ph II (ADAPT) neoadjuvant 12 weeks HER2+ HR+ BC
Kadcyla + Perjeta: Ph III (MARIANNE) in 1L HER2+ mBC
Herceptin + Perjeta: Ph II (NEOPSPHERE) in neoadjuvant HER2+ BC
Approved in US, filed in EU
Avastin + Letrozole: Ph III (CALGB40503) in 1L HR+ mBC
Hematology
Gazyva: Ph III (GADOLIN) in R/R iNHL
Filing 2015
Venetoclax: Ph I in R/R NHL and R/R MM
BTD: Filing 2015 in 17p CLL depleted
92
Cancer immunotherapy: Encouraging early data
in monotherapy across various cancer types
Range across various studies
NSCLC 2L
ORR range
mPFS range
mOS* range
Prevalence
Un-stratified
~ 15%
~3m
~ 9-11 m
Stratified
40-45%
~ 6-8 m
Not reached
~20-30%
ORR range
mPFS range
mOS range
Prevalence
25-45%
~ 2-5.5 m
Not reached
~30%
Range across various studies
Bladder 2L
Stratified
* Comparisons based on publicly available competitor data sets. Significant limitations with these cross-trial comparisons; significant variability
in baseline population demographics further limit conclusions
Note: CheckMate 057 2L+ non-squamous NSCLC trial was stopped early because the study met its OS endpoint
93
Cancer immunotherapy: Still many open questions
Immunity
Review
T Cell Trafficking
What are the factors affecting
T-cell infiltration?
Antigen Presentation
Biology
How can we convert unresponsive
tumors into responsive tumors?
Immunosuppression
Can we modulate / monitor T-cell
responses?
Figure 1. The Cancer-Immunity Cycle
Biomarker/
Tools
The generation of immunity to cancer is a cyclic process that can be self propagating, leading to an accumulation of immune-stimulatory factors that in principle
should amplify and broaden T cell responses. The cycle is also characterized by inhibitory factors that lead to immune regulatory feedback mechanisms, which
can halt the development or limit the immunity. This cycle can be divided into seven major steps, starting with the release of antigens from the cancer cell and
ending with the killing of cancer cells. Each step is described above, with the primary cell types involved and the anatomic location of the activity listed. Abbreviations are as follows: APCs, antigen presenting cells; CTLs, cytotoxic T lymphocytes.
Can we create a personalized immunotherapy paradigm?
to (step 4) and infiltrate the tumor bed (step 5), specifically recognize and bind to cancer cells through the interaction between its
T cell receptor (TCR) and its cognate antigen bound to MHCI
(step 6), and kill their target cancer cell (step 7). Killing of the cancer cell releases additional tumor-associated antigens (step 1
again) to increase the breadth and depth of the response in subsequent revolutions of the cycle. In cancer patients, the CancerImmunity Cycle does not perform optimally. Tumor antigens may
not be detected, DCs and T cells may treat antigens as self rather
than foreign thereby creating T regulatory cell responses rather
than effector responses, T cells may not properly home to
tumors, may be inhibited from infiltrating the tumor, or (most
importantly) factors in the tumor microenvironment might suppress those effector cells that are produced (reviewed by Motz
and Coukos, 2013).
The goal of cancer immunotherapy is to initiate or reinitiate a
self-sustaining cycle of cancer immunity, enabling it to amplify
and propagate, but not so much as to generate unrestrained
autoimmune inflammatory responses. Cancer immunotherapies
dampen or arrest the antitumor immune response, the most
effective approaches will involve selectively targeting the ratelimiting step in any given patient. Amplifying the entire cycle
may provide anticancer activity but at the potential cost of
unwanted damage to normal cells and tissues. Many recent clinical results suggest that a common rate-limiting step is the immunostat function, immunosuppression that occurs in the tumor
microenvironment (Predina et al., 2013; Wang et al., 2013).
Initiating Anticancer Immunity: Antigen Release,
Presentation, and T Cell Priming
Attempts to activate or introduce cancer antigen-specific T cells,
as well as stimulate the proliferation of these cells over the last 20
years, have led to mostly no, minimal or modest appreciable
anticancer immune responses. The majority of these efforts
involved the use of therapeutic vaccines because vaccines can
be easy to deploy and have historically represented an approach
that has brought enormous medical benefit (reviewed by Palucka and Banchereau, 2013). Yet, cancer vaccines were limited
on two accounts. First, until recently, there was a general lack of
Indication/
Patient pop.
Can we convert an unresponsive tumor to a responsive one?
Therapy
Efficacy/ Safety
What influences durability of response?
94
Doing now what patients need next