0 2 4 4 8 7 2 9 6 1 2 0 1 4 4 1 6 8 1 9 20 .0 1

0 .1

1

1 0

1 0 0

1 0 0 0

P la s m a C o n c e n tra tio n

H o u rs P o s t-D o s e

nM

a c t-T C B , 1 8 0

a c t-T C B , 6 0

P b -T C B , 2 0 0 0

A164

INTRODUCTION

CONCLUSIONS

RESULTS

T cell-engaging bispecific antibodies (TCBs) are highly potent therapeutics which direct the activity of cytotoxic T cells to tumors. TCBs have shown clinical activity in hematologic malignancies, but development of TCBs for solid tumor indications is proving more challenging. Due to their high potency, TCBs can target normal tissues with low antigen expression, resulting in significant on-target, off-tumor toxicity that can limit dosing to low levels. As a result, it has been difficult to reach the level of drug exposure required for efficacy without excessive toxicity. Therefore, novel methods are needed to enable the potent anti-tumor activity of TCBs while minimizing toxicity due to cytokine release and damage to healthy tissues. CytomX has developed a new class of recombinant, proteolytically-activated antibody prodrugs (ProbodyTM therapeutics) that are “masked” to reduce binding to antigen in healthy tissue, but can become “unmasked” by proteases that are preferentially activated in the tumor microenvironment. In this way, Probody therapeutics are designed to increase therapeutic index by maximizing efficacy and minimizing on-target toxicity in normal tissues. Here we describe a T cell-engaging Bispecific Probody therapeutic (Pb-TCB) targeting Epidermal Growth Factor Receptor (EGFR) and CD3 that has been optimized for affinity, effector function, masking, and cleavability. In vitro, under protease-deficient conditions, we demonstrate that the unmasked EGFR-CD3 TCB has potent, EGFR-dependent tumor cell killing, while the doubly-masked EGFR-CD3 Pb-TCB reduces target-dependent cytotoxicity by more than 100,000-fold. However, in established tumor models where tumor-resident proteases are expected to be active, we demonstrate that Pb-TCBs potently induce tumor regressions. Further, in non-human primates, the maximum tolerated dose (MTD) of the EGFR-CD3 Pb-TCB is more than 60-fold higher than the MTD of the unmasked TCB, and the tolerated exposure (AUC) is more than 10,000-fold higher. Finally, even at a 60-fold higher dose, transient serum cytokine and AST/ALT increases observed in non-human primates treated with the Pb-TCB are still lower than those induced by the TCB. By localizing activity to the tumor microenvironment, Pb-TCBs have the potential to expand clinical opportunities for T cell-engaging bispecific therapies that are limited by on target toxicities, especially in solid tumors. Moreover, an EGFR-CD3 Pb-TCB has the potential to address EGFR-expressing tumors that are poorly responsive to existing EGFR-directed therapies.

PROBODY is a trademark of CytomX Therapeutics, Inc. All other brands and trademarks referenced herein are the property of their respective owners.

EGFR-CD3 Bispecific Probody™ Therapeutic Induces Tumor Regressions and Increases Maximum Tolerated Dose >60 fold in Preclinical Studies Leila M. Boustany, Sherry L. La Porte*, Laurie Wong, Clayton W. White, Linnea Diep, Yuanhui Huang, Shouchun Liu, Jennifer H. Richardson, W. Michael Kavanaugh, Bryan A. Irving* � CytomX Therapeutics, Inc., South San Francisco, CA

Figure 2: CytomX Probody T Cell-Engaging Bispecific (Pb-TCB) Format

Binding of the Pb-TCB and act-TCB to EGFR+ HT29 cells (A) and CD3+ Jurkat cells (B) was evaluated by flow cytometry. Masking efficiency (shift in apparent Kd of Pb-TCB relative to act-TCB) of the EGFR mask is approx. 1000 and is >5000 for the CD3 mask.

• EGFR/CD3 act-TCB demonstrates potent EGFR-dependent T cell mediated cell killing at sub pM concentrations in vitro.

• The Probody-TCB attenuates EGFR and CD3 on cell binding and reduces targeted, T cell mediated cytotoxicity by over 300,000-fold in vitro.

• Pb-TCB and act-TCB induce regressions of established EGFR+ tumors in human PBMC engrafted NSG mice.

• Substrate cleavability correlates with both efficacy and tumor T cell infiltration in PBMC engrafted NSG mice.

• In cynomolgus monkeys, masking increases the maximum tolerated dose by 60-fold and reduces both cytokine release and measures of acute organ toxicity relative to the unmasked TCB.

• In cynomolgus monkeys, masking enables sustained plasma exposure. • In preclinical studies, potent anti-tumor efficacy combined with a striking improvement in

safety demonstrates the advantage of Probody technology for T cell-engaging bispecific therapies.

Figure 4: Pb-TCB Offers Substantial Protection in in vitro Functional Assays

Figure 6: EGFR/CD3 Pb-TCB Induces Regressions of Established HT-29 Tumors in PBMC Engrafted NSG mice

Female NSG mice (n=8/group) were implanted SC with 2 million HT29Luc2 cells on day -15. Three days later, mice were injected IP with human PBMCs at a T cell/tumor inoculum ratio of 1:1. Test and control articles were administered at 0.3 mg/kg, weekly IV as shown in the above schematic. *NSUB is a Pb-TCB devoid of a cleavable substrate sequence (No-SUBstrate). TV is presented as mean ±SEM.

Figure 3: Pb-TCB Demonstrates Reduced Binding to EGFR+ and CD3+ Cells in vitro

Figure 7: EGFR/CD3 Pb-TCB is Efficacious in HCT116 Established Tumor Model

A.  Pb-TCB shifts EC50 of T cell mediated cell killing >300,000x. No cytotoxicity is observed when a non-EGFR binding Pb-TCB is used demonstrating that target engagement is required for activity. Cytotoxicity was quantified using Promega OneGlo assay following a 48 hour incubation of PBMCs and HT29 Luc2 target cells at a 10:1 ratio. Co-cultures were treated with Pb-TCB or act-TCB at the concentrations shown above. B.  The Pb-TCB demonstrates reduced T cell activation relative to the

unmasked TCB. The dose response curve for CD8+ T cell activation is shifted for the Pb-TCB relative to the act-TCB. A ~300x higher concentration of the masked molecule is required to activate T cells relative to the unmasked molecule. CD8+ T cell activation was measured by flow cytometry as the percentage of CD69+CD8+ T cells following a 48 hour incubation of PBMCs and U266 cells at a 10:1 ratio. Co-cultures were treated with Pb-TCB or act-TCB at the concentrations shown above.

Figure 8: Pb-TCB Provides Increased Safety Relative to act-TCB in Cynomolgus Monkeys

B Pb-TCB shifts dose-response for cytokine release and T cell activation relative to act-TCB

Cytokine analysis was performed with a Luminex® suspension array system on serum samples. Data presented were obtained at 8 hours post-dose. Flow cytometry was performed on permeabilized samples. Data presented were obtained at 72 hours post-dose.

C Pb-TCB shifts dose response for acute liver toxicity relative to act-TCB.

Serum chemistry was assessed pre-dose, 48 and 168 hours post-dose. Data presented were acquired 48 hours post-dose.

day -15 HT-29Lu2 SC

day -12 hPBMC IP

day 0 day 7 day 14

day 21 Dose:

Figure 5: Pb-TCB Sensitivity to Protease Cleavage Correlates with Tumor T cell Infiltration and Efficacy in PBMC Engrafted NSG Mice

Figure 9: Tolerated Pb-TCB Exposure is > 10,000-fold Higher than Tolerated Exposure of act-TCB in Cynomolgus Monkeys

Cleavage by tumor associated proteases

α-CD3 scFvs

α-EGFR masks

Protease substrate

Fc effector mutant

α-EGFR

•  Full IgG bispecific format to maximize exposure and half-life •  Fc-effector impaired to minimize cross linking to FcγR bearing cells •  Format optimized for a-CD3 affinity, mask strength and cleavable substrates •  act-TCB represents protease activated, unmasked TCB

α-CD3 scFvs

Parental Antibody

Substrate Linker - Stable in vivo in circulation - Preferentially cleaved by tumor proteases

Figure 1: Probody Therapeutics are Protease- Activatable Antibody Pro-Drugs

Masking Peptide -  Blocks antigen binding -  Released from Probody therapeutic upon substrate cleavage, producing a fully active antibody

Protease - Dysregulated activity of specific proteases in cancer tissue - Cleaves mask to enable Probody therapeutic binding

1 0 -3 1 0 -2 1 0 -1 1 0 0 1 0 1 1 0 2 1 0 3 1 0 4

0

5 0 0 0

1 0 0 0 0

1 5 0 0 0

2 0 0 0 0

2 5 0 0 0

C D 3 B in d in g (J u rk a t)

[A c t/P b -T C B ] n M

MF

I (c

orr

ec

ted

)

P b -T C B

a c t-T C B

1 0 -3 1 0 -2 1 0 -1 1 0 0 1 0 1 1 0 2 1 0 3 1 0 40

5 0 0 0

1 0 0 0 0

1 5 0 0 0

2 0 0 0 0

2 5 0 0 0

3 0 0 0 0

3 5 0 0 0

4 0 0 0 0

E G F R B in d in g (H T 2 9 )

[A c t/P b -T C B ] n M

MF

I (c

orr

ec

ted

)

P b -T C B

a c t-T C B

A B

1 0 -4 1 0 -3 1 0 -2 1 0 -1 1 0 0 1 0 1 1 0 2 1 0 3 1 0 4 1 0 5 1 0 6

0

2 0

4 0

6 0

8 0

1 0 0

H T 2 9 C y to to x ic ity

[A c t/P b -T C B ] p M

% C

yto

tox

icit

y

P b -T C B

n o n E G F R b in d in g P b -T C B

a c t-T C B

1 0 0 1 0 1 1 0 2 1 0 3 1 0 4 1 0 5 1 0 6 1 0 70

2 0

4 0

6 0

8 0

1 0 0

C D 8 T c e ll A c t iv a t io n

[A c t/P b -T C B ] p M

CD

69

(%

po

sit

ive

)

P b -TC B

a c t-TC B

UT

360,000x

A B

Pb-TCB with higher protease sensitivity leads to greater T cell infiltration

CD3+ cells (brown staining) in HT29 tumors harvested 7 days after a 1 mg/kg dose

20X mag

NSUB* Pb-TCB 2 Pb-TCB 1 act-TCB PBS

(Substrate Cleavability: Pb-TCB 2> Pb-TCB 1> NSUB*)

A

0 5 1 0 1 5 2 00

2 0 0

4 0 0

6 0 0

H T 2 9 X e n o g ra ft

S tu d y D a y

Tu

mo

r V

olu

me

(m

m3)

N SUB *

P b -TC B 2

P B S

a c t-TC B

P b -TC B 1

B Pb-TCB with higher protease sensitivity leads to greater efficacy

0 5 1 0 1 5 2 0 2 50

2 0 0

4 0 0

6 0 0

8 0 0

S tu d y D a y

Tu

mo

r V

olu

me

(m

m3)

H T 2 9 X e n o g ra ft

P B S

P b -T C B , 1 .5

P b -T C B , 0 .5

* * * *

*

Female NSG mice (n=7/group) were implanted SC with 2 million HT29Luc2 cells on day -15. Three days later, mice were injected IP with human PBMCs at a T cell/tumor inoculum ratio of 1:1. Test and control articles were administered at 0.5 or 1.5 mg/kg, weekly IV. TV is presented as mean ±SEM.

0 5 1 0 1 5 2 0 2 50

2 0 0

4 0 0

6 0 0

8 0 0

1 0 0 0

1 2 0 0

H C T 1 1 6 X e n o g ra ft

S tu d y D a y

Tu

mo

r V

olu

me

(m

m3)

P b -T C B , 0 .3

P b -T C B , 1

P B S

a c t-T C B , 0 .3

**

********

Female NSG mice (n=8/group) were implanted SC with 2 million HCT116 cells on day -15. Three days later, mice were injected IP with human PBMCs at a T cell/tumor inoculum ratio of 1:1. Test and control articles were administered IV at 0.3 mg/kg or 1 mg/kg, weekly. TV is presented as mean ±SEM.

TCB Dose(µg/kg) ClinicalObserva9onsact-TCB

60(MTD)

Moderate

act-TCB 180 Severe

Pb-TCB 600 None

Pb-TCB 2000 Mild

Pb-TCB

4000(MTD)

Moderate

Pb-TCB 6000 Severe

A Pb-TCB has >60 fold higher MTD than act-TCB

1 0 1 0 0 1 0 0 0 1 0 0 0 00

1 0 0 0

2 0 0 0

3 0 0 0

4 0 0 0

IF N g

D o s e (mg /k g )

pg

/ml

1 0 1 0 0 1 0 0 0 1 0 0 0 00

2 0

4 0

6 0

8 0

K i6 7 + C D 4 +

D o s e (mg /k g )

% K

i-6

7+

CD

4+

1 0 1 0 0 1 0 0 0 1 0 0 0 00

1 0 0 0 0 0

2 0 0 0 0 0

3 0 0 0 0 0

4 0 0 0 0 0

IL -6

D o s e (mg /k g )

pg

/ml

a c t-T C BP b -T C B

1 0 1 0 0 1 0 0 0 1 0 0 0 00

2 0 0

4 0 0

6 0 0

8 0 0

A S T

D o s e (mg /k g )

U/L

a c t-TC B

P b -TC B

1 0 1 0 0 1 0 0 0 1 0 0 0 00

5 0

1 0 0

1 5 0

2 0 0

2 5 0

A L T

D o s e (mg /k g )

U/L

Plasma concentration of act-TCB and Pb-TCB was measured by ELISA using anti-id capture and anti-huFc detection. Time points after 4 hours for act-TCB dosed at 60 µg/kg and 24 hours for act-TCB dosed at 180 µg/kg were BLQ. Tolerated exposure represents area under the curve (AUC) of Pb-TCB (448 day*nM) dosed at 2000 µg/kg and act-TCB (0.04 day*nM) dosed at 60 µg/kg.

Pb-TCB act-TCB

© 2017, CytomX Therapeutics, Inc.

* Affiliation at the time of the study