The Next Era of Cancer Therapeutics: Defining Biologic Problems and Engineering Solutions in Cancer Immunotherapy
PEGS Virtual MeetingWednesday June 3rd, 2020
Daniel S. Chen MD PhDChief Medical OfficerIGM Biosciences
DISCLOSURES
I am an employee of IGM Biosciences
The information presented is solely intended to foster the exchange of scientific and medical information.
CANCER IMMUNOTHERAPY IS A BREAKTHROUGH
Checkmate 214 OS(Ipi+Nivo 1L Int/Poor risk RCC)
Escudier, et al. ESMO 2017
Keynote 189 OS (CarboPem+Pembrolizumab 1L NSCLC)
Gandhi et al. NEJM 2018
PACIFIC PFS (Durvalumab Stage IIIB NSCLC)
Antonia, et al. N Engl J Med 2017
Horn et. al. NEJM 2018
IMpower133 OS Atezolizumab+chemo
1L Small Cell Lung
IMpassion130 OSAtezolizumab+chemo1L TNBC PDL1≥1%
IMpower150 PFSAtezolizumab+Avastin+chemo
1L NSCLC
HR, 0.617 P < 0.0001
Reck, et al. ESMO IO 2017
IMbrave150 OSAtezolizumab+Avastin
1L HCCHR, 0.58 P <0.001
Finn, et al. NEJM 2020Keynote 24 OS
(Pembrolizumab 1L TPS≥50% NSCLC)
Reck, et al. N Engl J Med 2016
HR, 0.70 P =0.0069
HR, 0.62 (95% CI 0.45-0.86)
Powles et. al. Lancet 2018
Chen DS PEGS Virtual 2020
CANCER IMMUNOTHERAPY IS A BREAKTHROUGH
Checkmate 214 OS(Ipi+Nivo 1L Int/Poor risk RCC)
Escudier, et al. ESMO 2017
Keynote 189 OS (CarboPem+Pembrolizumab 1L NSCLC)
Gandhi et al. NEJM 2018
PACIFIC PFS (Durvalumab Stage IIIB NSCLC)
Antonia, et al. N Engl J Med 2017
Horn et. al. NEJM 2018
IMpower133 OS Atezolizumab+chemo
1L Small Cell Lung
IMpassion130 OSAtezolizumab+chemo1L TNBC PDL1≥1%
IMpower150 PFSAtezolizumab+Avastin+chemo
1L NSCLC
HR, 0.617 P < 0.0001
Reck, et al. ESMO IO 2017
IMbrave150 OSAtezolizumab+Avastin
1L HCCHR, 0.58 P <0.001
Finn, et al. NEJM 2020Keynote 24 OS
(Pembrolizumab 1L TPS≥50% NSCLC)
Reck, et al. N Engl J Med 2016
HR, 0.70 P =0.0069
HR, 0.62 (95% CI 0.45-0.86)
Powles et. al. Lancet 2018
Chen DS PEGS Virtual 2020
Cancer immunotherapy is life altering… for some cancer patients
SO WHAT IS THE PROBLEM?
WHY CAN’T CANCER IMMUNOTHERAPY CURE CANCER MORE BROADLY?
A COMPLEX SET OF TUMOUR, HOST AND ENVIRONMENTAL FACTORS GOVERN STRENGTH, AND TIMING, OF ANTI-CANCER IMMUNE RESPONSES
Adapted from Chen and Mellman. Immunity 2013; Chen and Mellman. Nature 2017
∫(Fstim) - ∫(Finhib) ≥ 1/ n=1, y (TCRaffinity x frequency)Cancer immune set point:
Chen DS PEGS Virtual 2020
Cancer immune set pointCancer immunity cycle
TUMOR IMMUNITY CONTINUUM
IMMUNE DESERTCD8+ T cells are
absent from tumor and its
periphery
IMMUNE EXCLUDEDCD8+ T cells
accumulated but have not efficiently
infiltrated
INFLAMEDImmune cells
infiltrated, but inhibited
Inflamed Non-inflamed
Mechanisms associated with immune escape to PD-L1/PD-1 inhibition
Mutational Load, Endogenous retrovirus
TGF signalReactive stroma, MDSCs, Angiogenesis, Low MHC I
Respond favorably to PD-L1/PD-1 inhibition
IFNg signalPD-L1 expression, B-cells, TILs,
Intact Ag presentation
Fatty acid metabolismNeuroendocrine features
Wnt/b-cateninKi67hi,, Low MHC I
Adapted from Hegde et al., Clin Cancer Res, 2016.Hegde and Chen, Immunity 2020 Chen DS PEGS Virtual 2020
PROBLEM STATEMENT FOR CANCER IMMUNOTHERAPY
• Complex system: Immune system is highly complex with many positive and negative regulators comprising many cell types, circulating factors and compartments, all with spatial and temporal considerations
• Balance between anti-cancer immunity and autoimmunity: Simple systemic manipulation of immunity is often limited by stimulation of autoimmunity
• Cancer heterogeneity: Human cancer can be broadly separated into three phenotypes that feature three distinct mechanisms of immune escape with each likely driven by specific dominant biology
Chen DS PEGS Virtual 2020
HOW CAN ENGINEERED THERAPEUTICS HELP ADDRESS THESE CHALLENGES IN IMMUNOTHERAPY?
Chen DS PEGS Virtual 2020
1. ALLEVIATE IMMUNOSUPPRESSION
KILLCANCERCELLSInflamed
INFLAMEDCD8+ T
cells infiltrated,
but areinhibited
Sources: Modified from Chen DS, Mellman I. Immunity. 2013; Herbst et. al. Nature 2014; Hegde, Karanikas, Evers. Clin Cancer Res 2016
Chen DS PEGS Virtual 2020
ECHO-301/KEYNOTE-252: IDO INHIBITOR + PD1 INHIBITOR PHASE III STUDY
Long GV et al. ASCO 2018
MelanomaN~600
Epcadostat + pembrolizumab
Placebo + pembrolizumab
OS HR 1.13
Chen DS PEGS Virtual 2020
Community map of cancer immunity: A multitude of suppressive factors in Inflamed Cancer
https://cancer-immunity.nature.com/pages/map
Chen and Mellman. Nature 2017
• How many of the suppressive factors are important?
• How will we target multiple suppressive factors at once?
Chen DS PEGS Virtual 2020
2. AGONIZE AND ACTIVATE T CELLS ONLY IN THE TME
Sources: Modified from Chen DS, Mellman I. Immunity. 2013; Herbst et. al. Nature 2014; Hegde, Karanikas, Evers. Clin Cancer Res 2016
ACTIVATE Non‐inflamed
IMMUNE DESERT
Immune cells are absent from tumor
and its periphery
Chen DS PEGS Virtual 2020
IMMUNE DESERTS: A PROFOUND ABSENCE OF T CELLS AND OTHER IMMUNE CELLS IN THE TUMOR MICROENVIRONMENT
Passive:Potential lack of immunogenic antigens Lack of Antigen PresentationLack of appropriate co-stimulationImmune Hostile TMELack of immune attractive chemokines
Active:Repulsive Chemokines
Mutational Load , Endogenous retrovirus
Fatty acid metabolismNeuroendocrine features
Wnt/b-cateninKi67hi
Low MHC I
IMMUNE DESERTImmune cells are
absent from tumor and its
periphery
Adapted from Hegde et al., Clin Cancer Res, 2016.Hegde and Chen, Immunity 2019 publication pending
Possible MOAs?
Chen DS PEGS Virtual 2020
3. TARGETING IMMUNE EXCLUSION
RECRUIT/INFILTRATE
Excluded
EXCLUDED
CD8+ T cells present but
have not efficiently infiltrated
Sources: Modified from Chen DS, Mellman I. Immunity. 2013; Herbst et. al. Nature 2014; Hegde, Karanikas, Evers. Clin Cancer Res 2016
Chen DS PEGS Virtual 2020
One of TGF’s functions is to trigger the formation of collagen fibers that can trap T cells:Will inhibiting TGF relieve this immune inhibition?
Bladder “excluded” phenotype: CD8/Collagen
Mariathasan, et al. Nature 2018
15
Chen DS PEGS Virtual 2020
M7824:Anti-PDL1/TGFb Trap
Strauss, et al. CCR 2018
TARGETING DOMINANT IMMUNE BIOLOGY IN THE TME
Chen DS PEGS Virtual 2020
ALTERNATIVE: BYPASS REGULATION OF ENDOGENOUS IMMUNITY
EXCLUDED
CD8+ T cells present but
have not efficiently infiltrated
INFLAMED
CD8+ T cells infiltrated,
but areinhibited
IMMUNE DESERT
Immune cells are absent from tumor
and its periphery
Sources: Modified from Chen DS, Mellman I. Immunity. 2013; Herbst et. al. Nature 2014; Hegde, Karanikas, Evers. Clin Cancer Res 2016
Synthetic Immunity Inflamed and Non‐Inflamed
Chen DS PEGS Virtual 2020
SYNTHETIC IMMUNITYEnable a potent “synthetic” anti‐cancer immune response targeted at tumor‐associated antigens (TAA)
Scott et. al. 2017
CAR-T
Anti-CD3
Anti-TAA
T cell engagerEnables immune-mediated elimination of cancer cells even when pMHC
machinery is lost or downregulated
Hegde and Chen, Immunity 2019 publication pending Chen DS PEGS Virtual 2020
SYNTHETIC IMMUNITY AND THE CANCER IMMUNITY CYCLE
Adapted from Chen and Mellman, Immunity 2013Hegde and Chen, Immunity 2020 Chen DS PEGS Virtual 2020
NEW DATA FROM AMERICAN SOCIETY OF HEMATOLOGY (ASH19): BCMA-TARGETED T CELL ENGAGERS AND CAR-T
Patients treated at 10mg (n=9): ORR 89%, 44% sCR/CR (MRD negative); CRS ~100%1 patient with Gr3/Gr5 CRS
6→10 mg and 10 mg
CC-932692:1 BCMAxCD3
(BMS)
Costa, et. al. ASH 2019 Berdeja, et. al. ASH 2019
Patients treated at 150x106 cells (n=12): ORR 83%, 33% sCR/CR (MRD negative); CRS ~67%, Neurotox 25%; 1 patient with Gr5 CRS at 450x106 cells
bb21217BCMA CAR-T
(BlueBird/BMS)
SYNTHETIC IMMUNITY: CHALLENGES FOR CART
• Homing of infused cells to TME
• Infiltration into TME
• Killing of target cells• Target Expression• Suppression
• CART cell survival• CART cell proliferation• CRS
• Cell preparation/logistics• Patient selection
• Promote endogenous immunity
Adapted from Chen and Mellman, Immunity 2013Hegde and Chen, Immunity 2020 Chen DS PEGS Virtual 2020
SYNTHETIC IMMUNITY: CHALLENGES FOR T CELL ENGAGERS
Adapted from Chen and Mellman, Immunity 2013Hegde and Chen, Immunity 2020
• Killing of target cells• Target Expression• Suppression
• CRS management and limitations to optimal dosing
• Adequate host T cells
• Promote endogenous immunity
Chen DS PEGS Virtual 2020
T CELL ENGAGERS: ACTIVATION AND FUNCTION
Chen DS PEGS Virtual 2020
YYYY YY YY YYYYYY YY YYYY YYYYYYYYYY YY YY
YYYY
YYYY YY YYYY
Plates coated with anti-CD3 antibodies
MOST OF WHAT WE THINK WE KNOW ABOUT T CELL ENGAGERS COMES FROM ANTI-CD3 COATED PLATES
Chen DS PEGS Virtual 2020
YYYY YY YY YYYYYY YY YYYY YYYYYYYYYY YY YY
YYYY
YYYY YY YYYY
Plates coated with anti-CD3 antibodies
Plate Bound T Cell Activation
T cell
MOST OF WHAT WE THINK WE KNOW ABOUT T CELL ENGAGERS COMES FROM ANTI-CD3 COATED PLATES
Chen DS PEGS Virtual 2020
Plates coated with anti-CD3 antibodies
T Cell Signaling and Activation
YYYY YY YY YYYYYY YY YYYY YYYYYYYYYY YY YY
YYYY
YYYY YY YYYY
Plate Bound T Cell Activation
T cellT cell dependent
cytotoxicity
+++
Generalized cytokine release
MOST OF WHAT WE THINK WE KNOW ABOUT T CELL ENGAGERS COMES FROM ANTI-CD3 COATED PLATES
Chen DS PEGS Virtual 2020
Plates coated with anti-CD3 antibodies
T Cell Signaling and Activation
YYYY YY YY YYYYYY YY YYYY YYYYYYYYYY YY YY
YYYY
YYYY YY YYYY
Plate Bound T Cell Activation
T cellT cell dependent
cytotoxicity
+++
Generalized cytokine release
MOST OF WHAT WE THINK WE KNOW ABOUT T CELL ENGAGERS COMES FROM ANTI-CD3 COATED PLATES
A highly artificial system to have based our understanding of underlying T cell biology on; Easy studies to perform, but likely does not represent actual physiologic conditions (eg TCR engagement density and level of cross linking super physiologic)
Chen DS PEGS Virtual 2020
EXAMPLE OF A NEXT GENERATION PLATFORM:ENGINEERED IGM ANTIBODIES
High AvidityEnhanced binding, binding to difficult and/or rare targetsEnabled for modification with high specificity, high affinity binding domains Enables affinity-avidity matching
“Auto” forms engineered bispecificAccelerates innovation by enabling rapid therapeutic generationExtended range of ratios between binding and engaging moieties
Czajkowsky et al.
Conformational change upon bindingEnables specialized properties upon binding
Can be designed to drive forceful agonistic properties
Rigid and polarized disc shape confers biophysical
propertiesEnhanced control of T cell engagement
Altered kinetics and compartmentalizationFurther engineering for control
Example: Engineered IgM Antibody
Platform for multi-specific molecules and additional
engineering Enables construction of protein
based ”nano-machines”
Naturally occurring formatMay be important in minimizing immunogenicity
Chen DS PEGS Virtual 2020
IGM CD20xCD3 binds very strongly to Ramos lymphoma cells at 37°C
HIGH AFFINITY HIGH AVIDITY IGM-BASED CD20XCD3 (IGM-2323) T CELL ENGAGER: BINDING
Keyu Li, Dean Ng, Ramesh Baliga
Koff for IGM-2323: unmeasurable
IgM CD20xCD3
IgG CD20xCD3
Anti-CD20
Anti-CD3
Anti-CD20
Anti-CD3
Chen DS PEGS Virtual 2020
IGM CD20xCD3 is significantly more potent against rituximab-resistantRamos Lymphoma cells expressing low levels of CD20
HIGH AFFINITY HIGH AVIDITY IGM-BASED CD20XCD3 T CELL ENGAGER: T CELL DEPENDENT CYTOTOXICITY
Antibody (pM)
IgM CD20xCD3IgG CD20xCD3
Rituximab
Marigold Manlusoc, Keyu Li, Ramesh Baliga, Bruce Keyt Chen DS PEGS Virtual 2020
IGM: DISSOCIATION OF CYTOTOXICITY AND “CYTOKINE RELEASE”
Dissociation of cytotoxicity and lack of cytokine release replicated in human, murine and non-human primate studies; Phase I study FPI late 2019, target complete dose escalation 2020
Keyt PEGS 2017
Lower cytokine release profile in vitro compared to IgG CD20 x CD3 bispecific antibody
IL-6 IL-2
IFNγ TNFα
Chen DS PEGS Virtual 2020
CD20XCD3 IGM BISPECIFIC T CELL ENGAGER: IN VIVO NON-HUMAN PRIMATE (NHP) EFFICACY AND SAFETY
Immunohistochemistry in vivo to detect CD19 and CD20 positive B cells in non-human primate spleen and mesenteric lymph nodes from NHP treated with a CD20 x CD3 IgM
IL-6 IL-2
IFNγ TNFα
Peak plasma inflammatory cytokine levels in a NHP dose study following treatment with CD20 x CD3 IgM
NHP dosed up to 25mg/kg of CD20 x CD3 IgM: no identified clinical or pathologic AEs/Tox Chen DS PEGS Virtual 2020
BIOPHYSICAL PROPERTIES OF ENGINEERED CANCER IMMUNOTHERAPY CAN BE MODIFIED TO OPTIMIZE EFFECTS
Deeg et al. Nano Lett 2013
Zal, Zal & Gascoigne
Miguel Otero Ritter et al.
Modified fromCalvo and Izquierdo 2018 other modification
physical characteristics of a T cell engager
• Shape• Size• Rigidity• Flex regions• Polarization• Charge• Coating/
Chen DS PEGS Virtual 2020
TCR SIGNALING: CYTOTOXICITY VS CYTOKINE RELEASE
Faroudi et al. PNAS 2003;100:24:14145-14150
Cytotoxicity as a function of antigenic stimulus
Functional T cell signaling is not dichotomous Two different activation thresholds in CTLs, with lytic threshold being more sensitive
Chen DS PEGS Virtual 2020
SYNTHETIC IMMUNITY: T CELL SIGNALING AND FUNCTION CAN DIFFER
CD3T cell
Cancer Cell
Bispecific T cell : Target Engager
IgG
Bispecific T cell :Target Engager
Single Chain Units
CD20
CAR-T
T cell
Cancer Cell
CD20
CAR
T cell
Cancer Cell
T cell receptor : MHC Engagement
TCR
MHC
CD3T cell
Cancer Cell
Bispecific T cell : Target Engager IgM
CD3
CD20
CAR-T, Chimeric antigen receptor-T cellMHC, Major histocompatibility complex plus peptide TCR, T cell receptor
Co-stimulatory domain
~3-10 TCR-pMHC
interactions required for CD8 T cell mediated killing of a
cancer cell†
† Purbhoo et al. Nature Immunology 2004 Chen DS PEGS Virtual 2020
•Stoichiometry (interaction strength vs two targets/cells)
• eg Stronger binding to cancer cells, more physiologic signaling to T cells
•Spatial (biophysical effects, signal strength)
• eg multimeric clustering, engagement density, compartment control•Temporal (prolonged vs pulsatile modulation)
• eg stability of signals, cellular context, limit overstimulation
EXAMPLE PLATFORM TECHNOLOGY: ENGINEERED IGM AND IGA Focus on Control of:
Chen DS PEGS Virtual 2020
WE ARE IN THE MIDST OF A GLOBAL PANDEMIC UNLIKE ANYTHING MOST OF US HAVE EVER SEEN IN OUR LIVES…
The need for advancements in virology, immunology and therapeutics could not be greater
Problem Statement for Pandemic Respiratory VirusPrevent infection, inhibit viral spread, modulate immunity
• Challenges for Antibody Engineering:
• How to irreversibly neutralize virus?• How to block viral variants?• How to provide appropriate access to viral sites of infection?• How to limit inflammation but enhance anti-viral immunity?• How to reduce Antibody Dependent Enhancement risk?
• Immunity is highly complex, requiring optimal immunotherapy to deliver tuned and orchestrated biologic modulation within specific compartments and to specific cells- at just the right amount
• Advances in antibody engineering may readily enable therapies that can provide such modulation in a convenient and accessible format
• Such approaches are well suited to cancer immunotherapy, but may also be applied to many other medical applications
CONCLUSIONS
Illustrative
Hegde and Chen, manuscript in preparation
Inno
vatio
n
57 years 46 years 49 years 16 years
Surgery(1846)
Radiotherapy(1903)
Chemotherapy(1949)
Targeted therapy(1998)
Cancer Immunotherapy(2014)
Amplitude
Wavelength (l)
Time
Synthetic immunity (CAR-T, etc)(2017)
Chen DS PEGS Virtual 2020
ACKNOWLEDGEMENTS
• Angus Sinclair• Beatrice Wang• Tasnim Kothambawala• Ling Wang• Manal Amoury• Steve Carroll• Maya Leabman• Maya Kotturi• Marvin Peterson• Avneesh Saini• Madeline Tran
IGM Biosciences• Bruce Keyt• Kathy Miller• Paul Hinton• Dean Ng• Keyu Li• Kevin Carlin• Sachi Rahman• Yasinee Ng• Marigold Boe
• Wayne Godfrey• Eric Humke• Genevieve Hernandez• Iris Sison• Rachel Mejia
• Ramesh Baliga
• Fred Schwarzer • Misbah Tahir
Genentech/Roche• Pablo Umana• Ira Mellman• Priti Hegde• CITC
325 East Middlefield Road Mountain View, CA 94043 Chen DS PEGS Virtual 2020