Tumor Targeting of a STING Agonist by Means of an

Antibody-Drug Conjugate Induces Potent Anti-Tumor

Immune Responses

Marc Damelin, Ph.D.

Innate Immunity Stimulating Therapies Summit Digital Event – July 22, 2020

Innovative and Highly Differentiated ADC Technologies and Platforms

2DAR = Drug-to-antibody ratioSTING = Stimulator of Interferon Genes

Efficacy without severe neutropenia,

neuropathy, or ocular toxicity

DolaLock Dolaflexin

Improved therapeutic index vs. other

platforms

Dolasynthen

Homogenous & Customizable Platform

Immunosynthen

Systemic administration with targeted immuno-stimulatory effect

• Controlled bystander effect

• Selectively toxic to rapidly dividing cells

• Not a PgP substrate

• Induces immunogenic cell death

• DolaLock payload

• Polymer scaffold

• DAR ~10-12

• Excellent drug like properties

• DolaLock payload

• Synthetic scaffold

• Site-specific

• Precise DAR (2-24)

• Novel STING agonist

• Complete regression with one dose in multiple preclinical models

• Limited effect on systemic cytokines

• Preclinical evidence that STING activation induces prolonged anti-tumor activity and generates immune memory• Other agonists, including TLR7/8, have not shown similar activity in reported studies

• STING activation is more specific to potent Type I interferon gene activation, while TLR activation is associated with general inflammation

• Emerging clinical evidence that STING agonists (intratumoral injection) activate the pathway and do not have significant tolerability concerns

• STING agonists are highly compatible with bioconjugation through the platform technology as they have favorable physicochemical properties

• Oligonucleotides are less compatible (i.e. TLR9, RIG-1)

• Mersana has focused on non-CDN agonists

Why We Invested in STING over other Innate Immunity Pathways

3

ADCs are suited to overcome limitations of free agonist (intratumoral or IV)

– Targeted delivery reduces toxicity liabilities • Minimize toxicity to T and B cells by selective targeting of ADCs (T cell intrinsic function)• Minimize systemic inflammation

– Improved pharmacokinetics

– Accessibility to metastatic sites

– No restriction on tumor type, location or size

Strong Rationale for a STING ADC Approach

4

Holistic Approach to Build the Optimal STING ADC

5

1. Platform• Payload• Linker• Scaffold• Drug-to-Antibody Ratio (DAR)

2. Target and Antibody• Immune cell antigens• Tumor cell antigens• Tumor-associated antigens

Payload molecule

Drug load per scaffold

Charge balance

Aqueous solubilityBioconjugationAntibody

Evaluation included:• Analytical• In vitro characterization• In vivo characterization• Developability

• The target cell is not necessarily a tumor cell– Potential for new mechanisms for payload delivery– Implications for choice of targets and antibodies

• Optimal payload requirements are not known– Potency– Membrane permeability & efflux properties– Metabolism rate (once released)

• Special considerations for I-O in vivo studies– Xenograft models are grown in immune-compromised mice– Syngeneic models are not compatible with certain targets and antibodies

Immunostimulatory ADC Platform Development Cannot Be Based Solely on Cytotoxic ADC Experience

6

Platform Development

• Identified novel compounds representing multiple series– Compounds have a range of biological activity & diverse physicochemical properties

• Leveraged structure-based drug design (SBDD) and crystallography– Crystal structure solved with novel ligand bound to STING

• Filed IP

Novel STING Agonists Designed for ADCs

*Ligand has been removed from structure 8

0.01 1 100 100000

10000

20000

30000

40000

Log Drug Conc. (nM)

Benchmark1C3176C3165

IRF3

Act

ivat

ion

Examples ofnovelMersanapayloads

Payload 2

Payload 1

Leveraging Our ADC Expertise and Proprietary Technologies

10.0 12.5 15.0 17.5 20.0 2

ADCmAb

Aggregation

12.5 15.0 17.5 20.0 22

ADC (mAb #1)ADC (mAb #2)

Optimization

Aggregation in Initial STING ADCNo Aggregation after Optimization

(STING ADCs from 2 different mAbs)

Modular Synthemer Approach Enables ADC Optimization

Payload molecule

Drug load per scaffold

Charge balance

Aqueous solubilityBioconjugationAntibody

Summary of Exploratory Toxicology in NHP

10

• Evaluated ADCs based on 3 antibodies• Dosed up to 9 mg/kg antibody (~0.3 mg/kg STING agonist)• Repeat-dose and single-dose cohorts• Clinical observations

• All animals appeared normal throughout study• No changes in body temperature• No mortality or unscheduled euthanasia

• Toxicokinetics• High exposure after both administrations; dose dependent; overall profile similar to non-STING ADCs• ADC highly stable in circulation; minimal free payload in plasma

• Serum Cytokines• Transient, modest elevation of 5 cytokines out of 24 tested; similar to results in mouse

• No adverse changes in hematology or clinical chemistry• No adverse findings in histopathology

Targets and Antibodies

Comprehensive Approach to Target Validation and Selection

12Image based on DOI: 10.5772/intechopen.72648

Cell Types in the Tumor Microenvironment

Macrophages

Dendritic cellsTumor Cell and Tumor-associatedAntigens

STING Agonist ADCs with Complementary Therapeutic Rationales Based on Antigens and Target Cells

13

PotentialActivation

Activation

Delivery to 2 Cell Types

Target Category Rationale

Immune Cell • Direct activation of immune cells

Tumor Cell• Delivery to tumor and immune cells• Tumor-targeted delivery

Tumor-Associated• Proximity of antigen to immune cells• Tumor-targeted delivery

0.01 0.1 1 10 100 10000

10000

20000

30000

40000

Agonist [nM]

Luci

fera

se A

ctiv

ity

0.01 0.1 1 10 100 10000

50000

100000

150000

Agonist [nM]

Luci

fera

se A

ctiv

ity

Targeted ADC Targeted ADC – Fc MutantControl ADCFree Payload

AntigenAntigenAntigen

FcγR

Antibody variantsTargetedAntigen and

FcγR binding

ControlNo antigen binding

Fc mutantNo FcγR binding

X X

Co-culture assay

Immune CellIRF3 reporter(Luciferase)

Cancer Cell

Recombinant antigen assay

Immune CellIRF3 reporter(Luciferase)

Antigen-coated plate

Tumor-Targeted ADC Activates STING in Immune Cells

FcγRFcγR

Legend

Tumor-Targeted STING Agonist ADC Induces Killing of Cancer Cells by PBMCs

15

Vehicle

Targeted ADC

Control ADC

Free payload

1

10100

100

10

100

10

Agonist [nM]

- Cancer cells have stable expression of a red fluorescent protein in the nucleus

- These cancer cells have minimal STING activity

PBMCs

Co-culture:Cancer cells & PBMCs

100-Fold Increased Potency of ADC over Free Agonist

T Cells are Dispensable for Cancer Cell KillingSupports the hypothesized mechanism of action

PBMCs

CD3

CD

14

CD3+50%

Enriched Monocytes (CD14+CD16+)

CD3+4%

CD16-depleted Monocytes(CD14+CD16-)

CD3+2%

Control

Payload-4 nM

ADC-4 nM

Payload-20 nM

ADC-20 nM

PBMCs + Cancer cells

Control

Payload-4 nM

ADC-4 nM

Payload-20 nM

ADC-20 nM

Enriched Monocytes + Cancer cells Control

Payload-4 nM

ADC-4 nM

Payload-20 nM

ADC-20 nM

CD16-depleted Monocytes+ Cancer cells

16

Sustained Tumor Regressions Induced by a Single Administration of STING ADC

0 7 14 21 28 35 42 49 56 63-20

-10

0

10

20Body Weights

Days on Study

Bod

y W

eigh

t Cha

nge

(%)

Mea

n +/

- SEM0 7 14 21 28 35 42 49 56 63

0

250

500

750

1000Tumor Growth Inhibition

Days on study

Mea

n Tu

mor

Vol

ume

(mm

3 )±SE

M

diABZI I.V. agonist (0 / 5 mg/kg)

Control ADC (1 / 0.04 mg/kg)Control ADC (3 / 0.12 mg/kg)

Targeted ADC (1 / 0.03 mg/kg)Targeted ADC (3 / 0.09 mg/kg)

Unconjugated Antibody (3 / 0 mg/kg)

• All groups dosed IV• Doses in (mAb [mg/kg] / Payload [mg/kg])

17

Vehicle

Legend

10 / 10 Tumor Free

Target-Dependent Immune Cell Infiltration and Cytokine Induction in Tumors

18

Murine cytokine expression(qPCR on FFPE samples)

• ADC single dose• Tumors harvested 12 or 72 hrs post dose

Vehicle

12 hrs.

72 hrs.

Targeted ADC Control ADC

CD45 Immunohistochemistry

0

5

10

15

20

25C C 0

0

0.4

0.8

1.2

1.6

0

2

4

6

12 H 12 H 12 H 72 H72 H72 H

VehicleTargeted ADCControl ADC

mCXCL10 mIL-6mIFNβ

Dramatically Lower Induction of Serum Cytokines in Mice by STING ADC Compared to Free STING Agonist

CXCL10 IL-6 CXCL1 MIG (CXCL9)

RANTES (CCL5)

pg/m

L

0 20 40 60 800

1000

2000

3000

4000

Hours

0 20 40 60 800

5000

10000

15000

20000

25000

Hours0 20 40 60 80

0

2000

4000

6000

Hours0 20 40 60 80

0

2000

4000

6000

Hours

0 20 40 60 800

5000

10000

15000

Hours

0 20 40 60 800

5000

10000

15000

Hours

0 20 40 60 800

1000

2000

3000

4000

5000

Hours0 20 40 60 80

0

50

100

150

200

Hours

0 20 40 60 800

20

40

60

80

Hours

0 20 40 60 800

200

400

600

Hours0 20 40 60 80

0

200

400

600

800

1000

Hours0 20 40 60 80

0

100

200

300

400

Hours

0 20 40 60 800

500

1000

1500

Hours

0 20 40 60 800

500

1000

1500

2000

Hours

0 20 40 60 800

20

40

60

80

Hours0 20 40 60 80

0

5

10

15

Hours

MIP-1a (CCL3)TNF-α IFNγ

pg/m

L

Luminex assay

0 5 10 1510

100

1000

10000

100000

Plasma Concentration

Time (Days)

Con

cent

ratio

n [n

g/m

l]M

ean

+/- S

D

Total AntibodyConjugated Drug

STING ADC Yields Extended Exposure

Vehicle control

Targeted ADC (0.09 mg/kg payload = dose for complete tumor regression)

Control ADC (0.09 mg/kg payload)

STING diABZI I.V. agonist (5 mg/kg payload = maximum tolerated dose; 37% tumor growth inhibition) 19

Another Target and Tumor Model:STING Agonist ADC Inhibits Tumor Growth After a Single Dose

20

0 7 14 21 28 35 42 49 56 63 70 77 840

300

600

900

1200Tumor Growth Inhibition

Days on Study

Mea

n Tu

mor

Vol

ume

(mm

3 )±SE

M

Control ADC (3 / 0.1 mg/kg)Targeted ADC (3 / 0.1 mg/kg)• All groups dosed IV• Doses in (mAb [mg/kg] / Payload [mg/kg])

Vehicle

Legend

0 7 14 21 28 35 42 49 56 63 70 77 84-20

-10

0

10

20

Body Weights

Days on study

Bod

y W

eigh

t Cha

nge

(%)

Mea

n±

SEM

Sustained Tumor Regressions After a Single Dose in a Syngeneic Model

0 7 14 21 28 350

500

1000

1500Tumor Growth - Group Means

Days on study

Mea

n Tu

mor

Vol

ume

(mm

3 )±SE

M

21

6/9 Tumor-free

2/10Tumor-free

Dose: mAb / payload (mg/kg)

0 7 14 21 28 35-20

-10

0

10

20Body Weight

Days on studyB

ody

Wei

ght C

hang

e (%

)M

ean±

SEM

0 7 14 21 28 35 42 490

500

1000

1500

2000

Targeted ADC (0.88 / 0.04 mg/kg)

Days on Study

Tum

or V

olum

e (m

m3 )

0 7 14 21 28 35 42 490

500

1000

1500

2000

Control ADC (1.09 / 0.04 mg/kg)

Days on studyTu

mor

Vol

ume

(mm

3 )0 7 14 21 28 35 42 49

0

500

1000

1500

2000Vehicle

Days on study

Tum

or V

olum

e (m

m3 )

Immunological Memory Induced by STING Agonist ADC

220 7 14 21 28

0

400

800

1200

1600Non-targeted tumor

Days post re-challenge

Mea

n Tu

mor

Vol

ume

(mm

3 )±SE

M UntreatedControls

6/6Tumors

• Tumor free mice re-implanted with:• Original targeted tumor on the opposite flank

(blue), and• Non-targeted tumor on the other flank (red).

• Untreated age matched mice also implanted as a control (black line).

Efficacy Study Rechallenge Study (Dual Flank)

0 7 14 21 280

50

100

150

200

250

Targeted tumor

Days post re-challenge

Mea

n Tu

mor

Vol

ume

(mm

3 )±SE

M

UntreatedControls

6/6Tumor-free

Dose: mAb/ payload mg/kg)

0 7 14 21 28 35 42 490

500

1000

1500

2000

Targeted ADC (0.88 / 0.04 mg/kg)

Days on Study

Tum

or V

olum

e (m

m3 )

6/9 Tumor-free

Re-challenge

Conclusions

23

1. STING agonist ADC platform• Novel agonist payload optimized for ADC

• Linker & scaffold designed to maximize therapeutic index

• Well-tolerated in non-human primates

2. In vivo activity in multiple targets, tumor models and mouse strains

3. Differentiation from IV agonist: activity and tolerability

4. Immunological memory

5. On track to nominate 1st Development Candidate in 2020

Mersana’s STING ADC Research Team

24

Discovery ChemistryJeremy DuvallJosh ThomasKeith BentleyBrian JonesEoin KelleherLiping YangRadha IyengarDorin Toader

BiologyRaghida Bukhalid Naniye CetinbasChen-Ni ChinScott CollinsTimothy EitasWinnie LeeTravis MonnellMarina ProtopopovaLiuLiang Qin Kelly SlocumMarc Damelin

Analytical & BioanalyticalKenneth AvocetienStephen BradleyTyler CarterSusan ClardyMark NazzaroElena Ter-OvanseyanLing XuJeffrey ZuritaDavid Lee

Translational MedicinePamela ShawRebecca Mosher

Chief Science & Technology OfficerTimothy B. Lowinger

BioconjugationKalli CatcottBingfan DuCheri StevensonAlex UttardAnouk DirksenDorin Toader