American Association for Cancer Research /March 29 — April 03, 2019 /Atlanta, GA

Poster #3256

RTX-224, an Allogeneic Red Cell Therapeutic Expressing 4-1BBL and IL-12, Exhibits Potent In Vitro and In Vivo Activity and a Favorable Preclincal Safety Profile

Anne-Sophie Dugast, Enping Hong, Maegan Hoover, Arjun Bollampalli, Douglas C. McLaughlin, Omkar Bhate, Timothy J. Lyford, Torben Straight Nissen, Christopher L. Carpenter, Thomas J. Wickham, Lori Melançon, and Sivan Elloul

Rubius Therapeutics, Cambridge, MA

RESULTS AND METHODSINTRODUCTION• Red Cell Therapeutics™ (RCTs) are a new class of allogeneic, off-the-shelf cellular therapy

product candidates for the treatment of cancer, rare diseases, and autoimmune diseases

• For the treatment of cancer, RCTs are engineered to recapitulate human immunobiology and activate immune responses by stimulating the adaptive and innate immune systems or induce a tumor-specific immune response by expanding tumor-specific T cells against a target antigen

• RCTs are engineered to express hundreds of thousands of copies of cell-surface costimulatory molecules and cytokines (Figure 1)

Figure 1. The RED PLATFORM® is Designed to Generate Allogeneic, Off-the-Shelf Cellular Therapies

RED PLATFORM®

CD34+ HEMATOPOIETIC

PRECURSOR CELLS

1 HealthyO-Negative

Donor

• The enucleated reticulocytes are Red Cell Therapeutics™ that express hundreds of thousands of biotherapeutic proteins on the cell surface

• Delivered at a dose of <1% of total red blood cell volume in the body

• Universal, scalable, and consistent manufacturing process

100-1000sof Doses

LENTIVIRAL VECTOR

ENCODING OF 4-1BBL & IL-12

EXPANSION &DIFFERENTIATION

ENUCLEATION& MATURATION

RED CELLTHERAPEUTICS

4-1BBL, 4-1 ligand; IL-12, interleukin-12.

Figure 2. RTX-224 Mediates Immunostimulatory Effects Through Simultaneous Presentation of the Costimulatory 4-1BBL and the Cytokine IL-12

4-1BBL

IL-12

4-1BBL

IL-12

T CellNK Cell

RTX-224 (4-1BBL+IL-12)Recapitulates Human Biology:Signal 2+3

CD4 CD8 NK

Th1 → IFNγProliferation

M2 → M1

IFNγProliferationCytotoxicity

CD56ProliferationCytotoxicity

AntigenPresentation

Antiangiogenic

IL-12

↑ IFNγ production↑ NK proliferation/activation↑ CD4 and CD8 proliferation/activation

4-1BBL

4-1BBL, 4-1BB ligand; IFNγ, interferon γ; IL-12, interleukin-12; M, macrophage; NK, natural killer; RCT, red cell therapeutic; Th1, type 1 T helper.

• RTX-224 is an allogeneic cellular therapy product candidate that simultaneously presents hundreds of thousands of copies of costimulatory 4-1BB ligand (4-1BBL) in its native form and the potent cytokine interleukin-12 (IL-12) on the cell surface to activate T and natural killer (NK) cells (Figure 2)

– 4-1BBL is a costimulatory molecule that can also drive T and NK cell proliferation and activation and interferon γ (IFNγ) production

– IL-12 is a pleiotropic cytokine that bridges innate and adaptive immunity by promoting type 1 T helper (Th1) immune responses, T and NK cell proliferation, CD8 T cell and NK cell cytotoxicity, and IFNγ production

• To date, the use of IL-12 and 4-1BB agonists for cancer immunotherapy has been limited because of systemic toxicities1,2

– Unlike recombinant cytokines and agonist antibodies, which are administered systemically, RCTs, such as RTX-224, are restricted to the vasculature, which may limit toxicities

OBJECTIVES• To assess the immunostimulatory and antitumor activity of RTX-224 in vitro and in vivo

• To determine if RTX-224 expressing 4-1BBL and IL-12 results in lower toxicity compared with systemic administration of 4-1BB agonist antibodies and recombinant IL-12

CONCLUSIONS

• In RTX-224, 4-1BBL and IL-12 have synergistic and complementary effects in promoting innate and adaptive immunity

• A murine surrogate of RTX-224 demonstrated potent antitumor activity in vivo in several tumor models when administered as monotherapy or in combination with a checkpoint inhibitors

– RTX-224 alone or in combination with an anti-PD-1 antibody promoted increased tumor regressions

• In contrast to systemic recombinant IL-12, RTX-224 resulted in no observed toxicity likely due to restriction of therapy to the vasculature

• RTX-224 demonstrated a potentially wide therapeutic window, with highly potent, preclinical antitumor activity and no observed toxicity

• Based on these results, clinical studies are planned to evaluate RTX-224 for the treatment of patients with solid tumors

REFERENCES1. Bartkowiak T, Curran MA. Front Oncol. 2015;5:117.

2. Lasek W, et al. Cancer Immunol Immunother. 2014;63:419-435.

ACKNOWLEDGMENTSMedical writing support was provided by Benjamin Levine, PhD, of Bio Connections LLC, sponsored by Rubius Therapeutics.

DISCLOSURESAll authors: employment with and equity ownership in Rubius Therapeutics.

Figure 3. RTX-224 Demonstrates Synergistic and Complementary Effects of 4-1BBL and IL-12 in Promoting the Expansion and Activation of Key Cell Subsets Driving Innate and Adaptive Immunity

SynergyNK Cell Count

Fold

Ch

ang

e

Fro

m M

ed

ia C

on

tro

l

15

10

5

0

MediaRTX-CTRL

RTX-IL-12

RTX-4-1BBLRTX-224

rIL-12

A

Synergy

IFNγIF

Nγ,

pg

/m

L

4 x 105

3 x 105

2 x 105

1 x 105

0

MediaRTX-CTRL

RTX-IL-12

RTX-4-1BBLRTX-224

rIL-12

B

4-1BBL DrivenCD4 T Cell Count

Fold

Ch

ang

e

Fro

m M

ed

ia C

on

tro

l

MediaRTX-CTRL

RTX-IL-12

RTX-4-1BBLRTX-224

rIL-12

5

4

3

21

0

C

4-1BBL and IL-12 DrivenCD8 T Cell Count

MediaRTX-CTRL

RTX-IL-12

RTX-4-1BBLRTX-224

rIL-12

10

8

6

42

0

Fold

Ch

ang

e

Fro

m M

ed

ia C

on

tro

l

D

IL-12 DrivenCytotoxicity

MediaUntreated

RTX-IL-12

RTX-4-1BBLRTX-224

100

80

60

4020

0

Targ

et

Ce

ll D

eat

h, %

E

RTX cells were incubated with human PBMCs at different ratios in the absence or presence of anti-CD3 for 5 or 8 days, respectively. (A, B) RTX cells were incubated with human PBMCs at a 4:1, 2:1 or 1:1 ratio in the absence of anti-CD3 for 8 days. (A) Total NK cell counts were quantified using flow cytometry and depicted as fold-change over media control. (B) IFNγ concentration in culture supernatants was determined using enzyme-linked immunoassay (ELISA). (C, D) RTX cells were incubated with PBMCs in the presence of anti-CD3 (0.5ug/mL) at a 4:1, 2:1 or 1:1 ratio for 5 days. CD4 T cell (C), and CD8 T cell (D) counts were quantified using flow cytometry and depicted as fold-change over media control. (E) RTX cells were incubated overnight with purified human NK cells at a 5:1 ratio. K562 targets were added at a 5:1 effector:target ratio, and target killing was evaluated by flow cytometry. Data represent the average of 2-3 PBMC and NK cell donors.

CTRL, control; IFNγ, interferon γ; IL-12, interleukin-12; mRBC, murine red blood cell; NK, natural killer; PBMC, peripheral blood mononuclear cell; r, recombinant.

• RTX-224 induced potent antitumor effects associated with innate and adaptive immune responses in vitro (Figure 3)

– Innate immune response: RTX-224 demonstrated synergistic effects of 4-1BBL and IL-12 on NK cell proliferation and IL-12–driven effects on NK cell cytotoxic activity

– Adaptive immune response: RTX-224 demonstrated synergistic effects of 4-1BBL and IL-12 on stimulation of IFNγ production and proliferation of CD4 and CD8 T cells in vitro

Figure 4. A Murine Surrogate for RTX-224, mRBC-224, Demonstrates Tumor Growth Inhibition as Monotherapy and in Combination With an Anti-PD-1 Antibody in a MC38 Colon Cancer Mouse Model

00 4 8 12 16

500

1000

1500

Days After Randomization

Tum

or V

olu

me

, mm

3

mRBC-CTRLmRBC-CTRL+aPD-1mRBC-224mRBC-224+aPD-1

**

**

**

**

**

**

** ****

MC38 Model

Tum

or V

olu

me

, mm

3

00 4 8 12

500

1000

1500

Days After Randomization

mRBC–IL-12

Regression <50 mm3

D11

3/11

Tum

or V

olu

me

, mm

3

00 4 128

500

1000

1500

Days After Randomization

mRBC-224

Regression <50 mm3

D11

5/11

Tum

or V

olu

me

, mm

3

00 84 12

500

1000

1500

Days After Randomization

mRBC–IL-12+aPD-1

Regression <50 mm3

D11

4/11

Tum

or V

olu

me

, mm

3

00 4 8 12

500

1000

1500

Days After Randomization

mRBC-224+aPD-1

Regression <50 mm3

D11

9/11

mRBC-CTRL

mRBC-CTRL + aPD-1

mRBC-224

mRBC-224+ aPD-1

mRBC-CTRL

mRBC-CTRL + aPD-1

mRBC-224

mRBC-224+ aPD-1

mRBC-CTRL

mRBC-CTRL + aPD-1

mRBC-224

mRBC-224+ aPD-1

mRBC-CTRL

mRBC-CTRL + aPD-1

mRBC-224

mRBC-224+ aPD-1

* * ***** * *

* * *

0

20

40

60

80

IFNγ

+ C

D4

T C

ell

s, %

0

10

20

30

40

50

IFNγ+

NK

Ce

lls,

%

A

Pharmacodynamics

*

0

10

20

30

40

-2

2

0

4

6

8

CD

8+

CD

45 C

ell

s, %

CD

8 T

Ce

ll: R

eg

ula

tory

T C

ell

Rat

io

C

Tum

or V

olu

me

, mm

3

00 4 128

500

1000

1500

Days After Randomization

Tum

or V

olu

me

, mm

3

00 4 128

500

1000

1500

Days After Randomization

mRBC-CTRL

mRBC-CTRL+aPD-1

B

C57Bl6 mice were inoculated with MC38 (1×105 cells, subcutaneous). When tumors reached an average volume of 50 to 80 mm3, mice were administered 1×109 mRBC-CTRL, mRBC–4-1BBL, mRBC–IL-12, or mRBC-224 cells intravenously within 3- to 4-day intervals (days 1, 4, and 8; indicated by arrows). Some mice were administered with 150 μg of aPD-1 on days 1, 4, and 8. (A) Tumor volume was monitored over time. (B) This graph represents individual tumor volumes and the number of regressions in mice treated with mRBC–IL-12 or mRBC-224 in the presence or absence of 150 μg aPD-1 was displayed. (C) Tumors from mRBC-treated mice were harvested on day 11 and percentage of CD8 T cells, CD8 T cell:regulatory T cell ratio, and percentage of functional (IFNγ+) CD4 and CD8 T cells were reported.

* P ≤ .01, ** P ≤ .01, *** P ≤ .001, **** P ≤ .0001.

aPD-1, anti-programmed death-1; CTRL, control; IFN, interferon; IL-12, interleukin-12; mRBC, murine red blood cell; NK, natural killer

• Murine RBC-224 (mRBC-224) inhibited tumor growth greater than mRBC-expressing 4-1BBL or IL-12 alone (Figure 4)

• mRBC-224 plus a programmed death-1 antibody (aPD-1) resulted in a significant increase in the number of tumor regressions versus mRBC-224 monotherapy

• mRBC-224 stimulated a Th1 response, increased infiltration of CD8 T cells to tumors, and increased the functionality of CD4 T cells and NK cells in tumors in a mouse model

Figure 5. mRBC-224 Displays Antitumor Activity in Lung Metastasis and Subcutaneous B16F10 Tumor Mouse Models

Lung Metastases Subcutaneous Model

Lu

ng

Me

tast

ase

s, n

200

150

100

mRBC-CTRL

mRBC-IL-12

mRBC-4-1BBL

mRBC-224

3H3

50

0

Tum

or V

olu

me

, mm

3

2000

1500

1000

500

00 2 64 8 10 12

A B

********

********

Days After Randomization

mRBC-CTRLmRBC-2243H3

**

*

(A) C57Bl6 mice were inoculated with B16-F10 melanoma cells (1×105 cells, intravenously). Mice were administered 1×109 mRBC-CTRL, mRBC–IL-12, mRBC–4-1BBL, or mRBC-224 cells intravenously or 50 μg of 3H3 (agonist m4-1BB antibody) intraperitoneally as control on days 1, 4, and 8 after tumor inoculation. On day 14, mice were euthanized, and the number of lung metastases was quantified by counting colonies on the lung surface. (B) C57Bl6 mice were inoculated with B16-F10 melanoma (1×105 cells, subcutaneous). When tumors reached an average volume of 50 to 80 mm3, mice were administered 1×109 mRBC-CTRL, mRBC-224 cells intravenously or 50 μg of 3H3 intraperitoneally as control on days 1, 4, and 8 (indicated by arrows). Tumor volume was monitored over time.

*** P ≤ .001, **** P ≤ .0001.

CTRL, control; IL, interleukin; m4-1BB, murine 4-1BB; mRBC, murine red blood cell.

• mRBC-224 demonstrated significant tumor growth inhibition in both lung metastases and subcutaneous B16F10 models (Figure 5)

Figure 6. In Contrast to Systemic Treatment With Recombinant IL-12, mRBC-224 Results in No Observed Organ Toxicity In Vivo Likely Due to Restriction of Therapy to the Vasculature

RBCsF

15

10

5

0

RB

C, m

illio

ns/

µL

Spleen WeightE

0.6

0.4

0.2

0.0

We

igh

t, g

mRBC-CTRL

mRBC-224 (1×109 )

mRBC-224 (3×108 )

mRBC-224 (1×108 ) rIL-12

mRBC-CTRL

mRBC-224 (1×109 )

mRBC-224 (3×108 )

mRBC-224 (1×108 )rIL-12

rIL-12 (4 µg)

mRBC-CTRL

mRBC-224(1×108)

mRBC-224(3×108)

mRBC-224(1×109)

IFNγC

1×105

1×104

1×103

1×102

1×101

1×100

IFNγ,

pg

/m

L

10 14 183 7

Days After Randomization

Liver WeightD

2.0

1.5

1.0

0.5

0.0

We

igh

t, g

mRBC-CTRL

mRBC-224 (1×109 )

mRBC-224 (3×108 )

mRBC-224 (1×108 )rIL-12

ALTB

3 7 10 14 18

rIL-12 (4 µg)

0

500

1000

ALT

, U/

L

Days After Randomization

mRBC-CTRL

mRBC-224(1×108)

mRBC-224(3×108)

mRBC-224(1×109)

Change in Body WeightA

-200 1 2 3 4 5 6 7 8 9 10 11

-15

-10

-5

0

5

10

15

Bo

dy

We

igh

t C

han

ge

, %

rIL-12 (4 µg) + 3H3 (50 µg)

rIL-12

mRBC-CTRL

mRBC-224(1×108)

mRBC-224(3×108)

mRBC-224(1×109)

Days After Randomization

a a

C57Bl6 mice were administered 1×109 mRBC-CTRL, mRBC-224 (1×109, 3×108, or 1×108) intravenously, 3H3 at 50 μg per mouse intraperitoneally, or rIL-12 at 4 μg per mouse intraperitoneally on days 0, 4, 8, and 11. On day 18, mice were euthanized to measure change in liver toxicity and IFNγ in serum and total counts of RBCs in blood. (A) Body weight change over time. Bars represent mean (SD). (B) Serum levels of ALT were measured over time as units of enzyme per liter of blood. (C) IFNγ levels were measured in serum over time. (D,E) Liver and spleen weights were measured on day 18. (F) Total count of RBCs was measured in blood on day 18. Each point within each graph represents an individual mouse (D-F).

aMice died due to dramatic weight loss. Remaining two mice treated with recombinant IL-12 and 3H3 were excluded from further evaluation.

3H3, 4-1BB agonist mAb; ALT, alanine transaminase; CTRL, control; IFNγ, interferon γ; mRBC, murine red blood cell; rIL-12, recombinant interleukin-12.

• Mice treated with mRBC-224 maintained body weight over time, whereas mice administered systemic recombinant IL-12, with or without 4-1BB agonist antibody, demonstrated significant loss of body weight versus controls (Figure 6)

• mRBC-treated mice had no evidence of liver toxicity, which was demonstrated by no increase in serum alanine transaminase and liver weight, compared to mice treated with systemic recombinant IL-12