MSC Biology Mechanism and Clinical Design

Robert Deans, Athersys

ISCT NA Regional Meeting Philadelphia

9.8.13

The statements and discussions contained in this presentation that are not historical facts constitute forward-looking statements, which can beidentified by the use of forward looking words such as “believes,” “expects,” “may,” “intends,” “anticipates,” “plans,” “estimates” and analogousor similar expressions intended to identify forward-looking statements. These forward-looking statements and estimates as to futureperformance, estimates as to future valuations and other statements contained herein regarding matters that are not historical facts, are onlypredictions, and that actual events or results may differ materially. We cannot assure or guarantee you that any future results described in thispresentation will be achieved, and actual results could vary materially from those reflected in such forward-looking statements.

Information contained in this presentation has been compiled from sources believed to be credible and reliable. However, we cannot guaranteesuch credibility and reliability. The forecasts and projections of events contained herein are based upon subjective valuations, analyses andpersonal opinions. This presentation shall not constitute an offer to sell or the solicitation of an offer to buy any securities. Such an offer orsolicitation, if made, will only be made pursuant to an offering memorandum and definitive subscription documents

Clinical Development Pipeline

2013

MultiStem®Immunological

IBD / Ulcerative ColitisHematopoietic Stem Cell Transplant / GVHD

Solid Organ TransplantDiabetes

CardiovascularAMI

Congestive Heart FailurePeripheral Artery Disease

NeurologicalStroke

Traumatic Brain InjuryMultiple SclerosisSpinal Cord Injury

OtherBone Allograft

5HT2c Agonists (obesity, other)

IMMUNOLOGICALCARDIOVASCULAR

NEUROLOGICAL

2013-4 data

2013 Data

FDA approved P2

FDA discussions re: P2/3 design

German P1 approval

2

Phases of Commercialization (Circa NOW)

Minimally manipulated products regulated under tissue transplant guidelines and bypass

pre-market clinical testing

In some Regulatory environments, market entry of autologous products generates

revenue stream and enables development of allogeneic therapy

Clinical Proof of Concept

Running on All Cylinders

• Translational community addresses need for (industry) executing well designed studies and securing broad open access data registries

• Promoting an educated patient base with access to treatment without having to go underground

• Bioprocessing industry realizes business opportunity and invests in technology to bring manufacturing costs down

• Pharma and Healthcare see validated business model and invest with capital and development experience

• Regulatory policy accelerates access and responds to global alternatives

• Universities develop training programs tailored to cell therapy and bring next generation scientists into medicine

What Impact Can We Have As A Community?

• Better understand and control PK/PD profiles for cell therapy

• Collectively solve downstream cell processing technology improvements

• Educate Regulatory agencies around potency assays and cell comparability assays

• Influence reimbursement decisions and bring product COGS under realistic healthcare ceilings

Enhancing Healing Through Multiple Mechanisms

IMMUNOMODULATION

INFLAMMATION

REDUCTION

NEUROPROTECTION, CYTOPROTECTION

ANGIO- / VASCULOGENESIS

CELLULAR REGENERATION / REPLACEMENT BY HOST

PROGENITORS

Certain Secreted Proteins

Cytokines / Chemokines

Growth Factors

Other Proteins

SHIFTING BALANCE IN REPAIR PROCESSES

MultiStem® cells express multiple therapeutic proteins to enhance healing and tissue

repair in multiple ways

8Multimodal mode of action compared to single mode of biologic/drug

Pharmacokinetics

Infused cells reside in reticuloendothelial system with a

minority trafficking to site of injury

MultiStem Homing in Mouse GvHD Model

MultiStem Intravenous Infusion in Mouse GvHD Model

6 hours post-i.v. 24 hours post-i.v.

10

Limited MultiStem Retention Over Time

Day 1 Day 2 Day 3 Day 4 Day 5 Day 6 Day 7

Model:• LAD ligation and brachiocephalic vein

injection of luciferase labeled rat MultiStem® (5x105)

• Bioluminesent Imaging over one week

11Confidential

Whole Body Imaging Provides Improved Mass Balance Accounting

Whole Body Imaging

Quantifying Biodistribution

3-D Rendering of Cell Distribution in Liver, Femur

MAPC found in marginal zone of white pulp in spleen

Cell Migration to Spinal Cord Injury

Cardiovascular Disease

5 sec 30 sec 60 sec Vessel patency

Delivery of MultiStem® in AMI patient

Rapid, efficient delivery of cells to ischemic area

Retention of cells in relevant region of heart

Well tolerated

MultiStem®: Acute Myocardial Infarction

Blockage causingischemia

Angioplasty toclear blockage

Delivery of MultiStem

Biodistribution Following Transarterial Delivery in Pig Ischemia Model

19Medicetty et al, Cell Transplantation in press

Injection sites

Transarterial catheter delivery of pig MultiStem® cells, 2 weeks, animal 132

MultiStem Increases Ejection Fraction and Reduces Scar Size in Pig AMI Model

Source: Wang et al, Circulation 2009

20

MultiStem®: Acute Myocardial Infarction

Phase 1 Clinical Study completed- open label, dose escalation

DesignAdministered through coronary artery 2-5 days after AMI

Multiple clinical sites (Cleveland Clinic, Henry Ford and others)

Evaluation of safety

ResultsClean safety profile over 4 month period following treatment

Heart function benefit also observed: ~25% relative improvement in LVEF relative to baseline,

improvements in wall motion, and LV end systolic volume

0.00

4.00

8.00

12.00

16.00

20.00

Reg 20M 50M 100M

D E

F%

21Mean ± SEMPatients with Baseline LVEF of >30 – <45 at time of MultiStem administration

Context Setting – Cell Therapy for AMILVEF Comparison at Early Time Point (3-6 Months)

2.51.3 1.1 1.9

4.1

8.7

0.43

12.6

10

0

2

4

6

8

10

12

14

REPAIR-AMI

∆ 4m

ASTAMI∆ 6m

Janssens∆ 4m

Osiris∆ 3m

ATHX-20MM

∆ 4m

ATHX-50MM

∆ 4m

ATHX-100MM

∆ 4m

ATHX-45-20MM

ATHX-45-50MM

ATHX-45-100MM

Difference between Treated and (Control ) Patients

Change in mean

LVEF from baseline

compared to

(control)

*

PhasePatientsControl

II204

Placebo

II97

SOC

II67

Placebo

I53

Placebo

I25

Registry

Baseline LVEF ≤ 45 %

22

Note: ATHX data presented here not statistically significant

MultiStem v. HSC v. MSCPhase I / II Cardiovascular Cell Therapy Studies

23

AthersysMultiStem

Amorcyte (Neostem)HSC/CD34+

OsirisMSC

MesoblastMPC

Indication AMIAllogeneic

AMIAutologous

AMIAllogeneic

CHFAllogeneic

Delivery Transarterial Intracoronary Intravenous Endocardial

Study parameters

25 patients3 dose groups (20, 50, 100 MM cells)

Registry control

31 patients3 dose groups (5, 10,

15 MM cells)Placebo control

53 patients3 dose groups (0.5,

1.6, 5.0 MPK)Placebo control

60 patients3 dose groups (25, 75,

150 MM cells)Placebo control

∆ LVEFBL→Xmo

treatment v. control

4 months5.7% (pooled)

9.0% (LVEFBL ≤ 45%)12.6% (LVEFBL ≤ 45%,

50MM dose)

3 months2.5% (pooled)

3 months1.9% (pooled)

3 months3.4% (pooled)

Otheroutcomes

Improve stroke volume, wall motion,

ESV

Improve perfusion (in higher dose group)

Improve FEV1 Improve clinical outcomes, perfusion

Confidential

Angiogenesis Potency Assay

• Pre-clinical models show neo-angiogenesis in parallel with recovery benefit

• Screening of conditioned media shows expression of angiogenic factors, increased when exposed to inflammatory environment

• Confirmed by in vivo tissue microarrays

VEGF, CXCL5, IL-8

Neo-Angiogenesis Correlates with Response to Ischemia

Wang et al, Circulation 2009

AMIPVD

CHF

PBS

MultiStem

0

200

400

600

800

1000

1200

2 million 5 million 10 million PBS

vess

els

/mm

2

Vessel Density

Minimum Levels of VEGF, CXCL5 and IL-8 Required for Induction of Angiogenesis

IL-8VEGF

0 pg/ml

250 pg/ml

1000 pg/ml

2000 pg/ml

5 pg/ml

10 pg/ml

0 pg/ml

25pg/ml

0 pg/ml

20 pg/ml

60 pg/ml

80 pg/ml

CXCL5

0

10

20

30

40

50

60

Ave

# o

f Tu

be

s Fo

rme

d P

er

Fie

ld

0

10

20

30

40

50

60

70

EBM EGM SFM SFCM ISO 0pg/ml

IL8

20pg/ml

IL8

60pg/ml

IL8

80pg/ml

IL8

100pg/ml

IL8

Ave

# o

f Tu

be

s Fo

rme

d P

er

Fie

ld

Pass-Fail Criteria: Knockdown and Add-back Sets Critical Threshold Required for Activity

Consistency of Angiogenic Factor Production

• Tested culture supernatants using in vivo matrigelangiogenesis assay

• Determined levels of secreted angiogenic factor in spent media at time of clinical harvest

Statistical pass fail criteria

MultiStem (MAPC) promotes More Stable Blood Vessels than MSC and mesoangioblasts (Mab) in vivo

Roobrouck et al, Stem Cells 2011

Undifferentiated MSC, mesoangioblasts (Mab) and MAPC mixed with matrigel and VEGF165/bFGFwere transplanted under the skin of nude mice. (D) At 21 days, matrigel plugs were removed and examined macro- and microscopically (1nd and 2nd row) (2.5x and 6.6x respectively)

Human MSC and Mab-containing Matrigelsharbored leaky vessels, indicated with white arrows, which could be also seen on H&E stained cross-sections (last row) (40x)

28

Peripheral Vascular Disease (PVD)• Chronic occlusion of the lower extremity arteries

• Loss of blood flow results in pain, ischemic lesions, sepsis and can lead to limb amputation

• Can range from mild ( intermittent claudication) to severe ( critical limb ischemia: 5-10% of patients)

• Globally, up to 12% of global population is affected by PVD. Direct and indirect cost of treating PVD estimated > $2 billion.

• Risk factors: age, hypercholesterolemia, diabetes, smoking, obesity

• Formation of collaterals around site of stenosis can offer protection against ischemia and help restore blood flow

• Unmet need: current therapies have severe side effects, counterindicated for heart failure, many patients are unfit candidates for revascularization

29

Proximal Hindlimb Adductor Muscle Perfusion andFlow Reserve After MultiStem (MAPC) Injection

Day

BL 1 3 21

Blo

od

Flo

w (

mL

/min

/g)

0.0

0.2

0.4

0.6

0.8

1.0

Control(PBS)

Mic

rovascu

lar

Flo

w R

eserv

e

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

MAPC

*

Perfusion Flow Reserve(day 21)

ControlPBSMAPC

Inject

*

*

* p<0.05 vs control and PBS

30

MultiStem treatment to increase transverse arterioles

31

Control day 21 Control day 21

MAPC day 21 MAPC day 21

Examination of Arteriolar networks and distribution

32

Day 1 DAY 21

Untreated Controls

MultiStem

Cell Therapy as Disruptive Technology in Inflammation and Immunity

Dose-Dependent Modulation in T-Cell Assay

34

Stimulation

Classic MLR – reflects indirect antigen presentation by macrophages and naïve response

CD3/CD28 stimulation – reflects TCR signalling (direct Ag)

Flu Ag in MLR – reflects memory cell stimulation

Cytokine (IL-2, 7, 15) stimulation –reflects homeostatic T cell expansion

Readout

Inhibition of T cell proliferation

Prevention of T cell activation (FACS)

Inhibition of cytokine secretion

MultiStem’s Dynamic Regulation of Activated T-Cells

Activated T Cells

13 chemokines

5 cytokines

>25 secreted factors

>10 secreted factors

>40 receptors

3 receptors

>40 intracellular factors

>10 intracellular factors

3 chemokines

3 cytokines

>10 receptors

>20 intracellular proteins

3 intracellular proteins

> 5-fold difference in expression

MultiStem

35

MultiStem Interferes with Trafficking of Activated Immune Cells (Leukocyte Extravasation)

CD15s

36

MultiStem associated with limiting cell surface expression of CD15s on inflammatory cells and e-selectin and adhesion receptors on endothelial cells,

and inhibits leukocyte extravasation in multiple ways

Leukocyte extravasation:• Contributes to inflammation

and tissue damage (e.g., in regions of ischemia)

• Occurs mainly in post-capillary venules (minimized hemodynamic shear forces)

• Includes several steps, including chemo-attraction, rolling adhesion, tight adhesion, (endothelial) transmigration

• Process halted whenever any of these steps is suppressed

Inhibition of Inflammatory Cell Extravasation in AMI, Stroke

MultiStem

0

5

10

15

20

25

30

35

40

45

50

PBS (n=5) MAPC (n=4)

neu

tro

ph

ils/h

igh

po

wer f

ield

Neutrophil count in infarcted hearts Reduction in neutrophil tissue elastase

Rat Ischemia Model - direct injection in infarct zone; sacrifice after 3 days

p=.005 PBS Multistem

37

Tissue Microarray Evaluation in Ischemic Stroke Model Confirms Role

in Leukocyte Extravasation

Ki67

CD

4

No

MAPC

1:4No M

APC

1:4

0

1

2

3

4

% K

I67

+V

ENo M

APC

1:4

0

1

2

3

4

% K

I67

+V

E

No M

APC

1:4

0

2

4

6

8

% K

I67

+V

E

No M

APC

1:4

0

2

4

6

8

% K

I67

+V

E

No M

APC

1:4

0

5

10

15

20

25

% K

I67

+V

E

No M

APC

1:4

0

5

10

15

20

25

% K

I67

+V

E

* * *

IL-2 IL-7 IL-15

IL-2 IL-7 IL-15

Teff

Treg

0

20

40

60

80

100

% S

up

pre

ssio

n

IL-2

Teff

Treg

0

20

40

60

80

100

% S

up

pre

ssio

n

IL-7

Teff

Treg

50

60

70

80

90

100

% S

up

pre

ssio

n

IL-15

IL-2 IL-7 IL-15

*NS

*

FoxP

3

CD25

Treg

Teff

MAPCs facilitate expansion of Tregs during homeostatic proliferation

Solid Organ Transplant

Liver, Lung

AMS50

Graft >d50 after spleen cell application from

„tolerant“ animal

Heterotopic Heart Transplant

Allogeneic heart is grafted to adominalartery and venous

circulation – keeps on beating

MAPCs

(2x)Myco-

phenolate

(d0-7)

Lewis rat

ACI rat

Heterotopic Cardiac Allotransplant Model

0 20 40 60 80 1000

20

40

60

80

100

rejection day

Gra

ft S

urv

ival

(%

)Donor-type MAPC

+ MPA

3rd party MAPC+ MPA

MPA alone

41Confidential

LEW to ACI model (secondary skin Tx)

d0 d7 d9 d13 d15 d20

Lew ACI

Result from nine long term survivors: no acceptance of LEW skin.

representative picture

Heart Accepted / Skin Rejected – No Systemic Tolerance

Remarks

Even when skin graft is rejected in allogeneic response, heart remains unaffected!

Heterotopic heart can be transplanted to naïve allogeneic recipient - and is accepted without drug!

42Confidential

Secondary Tx:

Heart from LT acceptance animals

into ACI animals

NO IMMUNO-SUPPRESSION

0 20 40 60 80 1000

20

40

60

80

100

rejection day

Gra

ft S

urv

ival

(%

)

2 x 5 Mio LEWISMAPC + MPA is

Control

CsA therapy

0 20 40 60 80 1000

20

40

60

80

100

rejection day

Gra

ft S

urv

ival

(%

)

ControlHeart from LEWISMAPC therapyHeart from CsA therapy

100%

40%

CsA Treated

MAPC Treated

foxP3-APC

CD

25-P

E

CD4CD25highfoxP3

0

2

4

6

8ns.

*

Treg

s (%

of

lym

ph

ocy

tes)

ctr CsA MAPC

primary hearts

Primary Tx:

Lewis heart into ACI

animals + CsA or MAPC

MultiStem Provides Distinctive Benefit in Secondary Transplant

43Confidential

MiSOT-I Study Protocol Review – Dose Escalation

Summary

• Understanding mechanistic pathways and building PK/PD profile in disease models is key to clinical proof of mechanism

• Societies increasingly engaged in standards setting for cell characterization and potency– ISCT, ARM, CIRM, MSC Connect

• Nomenclature, reference standards, characterization assays– FDA, FACT, USP, DIA, Toxicology Society

• ISCT can be a core strength in advancing adherent cell therapies– Commercialization Committee / MSC Scientific Committee– Commercialization Committee – institutional membership

• WFIRM, UC Davis, Univ Loughborough, Karolinska, Catapult, UnivPenn

Immunology, GVHD – Richard Maziarz, Oregon Health Sciences University

AMI – Marc Penn, Summa Healthcare

Critical Limb Ischemia – Johnathan Lindner, Oregon Health Sciences University

Solid Organ Transplant – Marc Dahlke, Univ Regensburg DE

thanks again