Investor science conference call from

EULAR 2012

Berlin, 8 June 2012

2

This presentation contains certain forward-looking statements. These forward-looking

statements may be identified by words such as ‘believes’, ‘expects’, ‘anticipates’, ‘projects’,

‘intends’, ‘should’, ‘seeks’, ‘estimates’, ‘future’ or similar expressions or by discussion of,

among other things, strategy, goals, plans or intentions. Various factors may cause actual

results to differ materially in the future from those reflected in forward-looking statements

contained in this presentation, among others:

1 pricing and product initiatives of competitors;

2 legislative and regulatory developments and economic conditions;

3 delay or inability in obtaining regulatory approvals or bringing products to market;

4 fluctuations in currency exchange rates and general financial market conditions;

5 uncertainties in the discovery, development or marketing of new products or new uses of existing products, including without limitation negative results of clinical trials or research projects, unexpected side-effects of pipeline or marketed products;

6 increased government pricing pressures;

7 interruptions in production;

8 loss of or inability to obtain adequate protection for intellectual property rights;

9 litigation;

10 loss of key executives or other employees; and

11 adverse publicity and news coverage.

Any statements regarding earnings per share growth is not a profit forecast and should not be interpreted

to mean that Roche’s earnings or earnings per share for this year or any subsequent period will

necessarily match or exceed the historical published earnings or earnings per share of Roche.

For marketed products discussed in this presentation, please see full prescribing information on our

website – www.roche.com

All mentioned trademarks are legally protected

Introduction

Dr. Karl Mahler, Head of Investor Relations, Roche

Investor events 2012

R&D and strategy updates

4

Key assets / newsflow

Medical Meeting /

Event

Date /

location

4 June 2012

Chicago, USA

ASCO (American Society of Clinical

Oncology)

T-DM1: EMILIA pretreated HER2+ mBC

Avastin: TML treatment through multiple

lines in mCRC, AURELIA platinum resistant

ovarian cancer

Actemra: ADACTA RA monotherapy head to

head versus adalimumab

8 June 2012

Berlin, Germany

EULAR (Annual European Congress of

Rheumatology )

4/5 September 2012

London, GB

Investor Day / Dinner with Management R&D strategy and pipeline

Growth opportunities

Innovation and efficiency

Roche’s Inflammation portfolio

Focus on autoimmune and respiratory diseases and targeted immunotherapy

Status as of March 31, 2012

Phase I Phase II

New

Molecular

Entities

Additional

indications

CRTH2 antag asthma RG7185

IL-6 MAb RA CHU

TSLPR MAb asthma RG7258

IL-17 MAb autoimmune diseas. RG7624

IL-17 Mab inflammatory diseas. RG4934

rontalizumab SLE RG7415

LT alpha MAb RA RG7416

M1 prime MAb asthma RG7449

etrolizumab (β7) ulcerative colitis RG7413

Actemra systemic sclerosis RG1569 MabThera ANCA assoc vascul RG105

Xolair chronic idiopath. urticaria RG3648

Actemra early RA RG1569

Actemra DMARD IR H2H RG1569

Actemra RA sc formulation RG1569

Suvenyl enthesopathy CHU

lebrikizumab severe asthma RG3637

Phase III

5

Investor conference call from EULAR 2012

Agenda 8 June 2012 17:30 CET

6

• Introduction

Dr. Karl Mahler, Head of Investor Relations, Roche

• Actemra franchise overview

Dr. Karsten Jung, Global Product Strategy Head for Immunology & Ophthalmology

• Rheumatoid arthritis monotherapy in real-life setting: Re-evaluating need and

options

Prof. Paul Emery, Director of Musculoskeletal Division, Leeds Institute of Molecular

Medicine, University of Leeds, UK

• ADACTA: A phase IV study evaluating the reduction in disease activity during

monotherapy treatment with either Actemra or adalimumab

Prof. Paul Emery, Director of Musculoskeletal Division, Leeds Institute of Molecular

Medicine, University of Leeds, UK

• Q&A

Moderator: Dr. Karl Mahler

• Total duration 1.00 hour

Actemra/RoActemra franchise overview

Dr. Karsten Jung, Global Product Strategy Head for Immunology & Ophthalmology, Roche

Actemra / RoActemra*

Rheumatoid Arthritis (RA) is the core of the franchise

8 8

Three

core

elements

of

Actemra’s

life

cycle

Rheumatoid arthritis as core indication

Differentiation vs. other medicines in RA

Expanding beyond rheumatoid arthritis

*RoActemra - referred to as Actemra outside Europe & throughout the remainder of this document

Rheumatoid arthritis as core indication

Actemra: substantial evidence of efficacy to support broad label in RA

9

• DMARD IR sBLA submitted to the FDA Dec. 2011

- 65,000 patient years safety data

• 1st subcutaneous RA study (SUMMACTA, qw) positive,

2nd study (BREVACTA, q2w) expected Q3 2012

• Early RA data (FUNCTION) expected Q3 2012

9

0

40

80

120

160

200

Q1 08 Q1 09 Q1 10 Q1 11 Q1 12

+46%1

Actemra global sales

• 184 mCHF Q1 2012

CHF m

OPTION N=623 MTX IR

DAS28 remission at 24 wks, 8mg/kg

AMBITION N=673 MTX naive

TOWARD N=1,220 DMARD IR

RADIATE N=599 anti-TNF IR

LITHE N=1,196 MTX IR, X-ray

27.5%

30.2%

30.1%

33.6%

33.3%

Data across all key RA populations

in mono & combination therapy

Phase III clinical studies

1 constant exchange rates

• 4,009 ACTEMRA patients in clinical trials

• Median duration was 3.6 years

• Total observation time was 12,293 patient-years

Rheumatoid arthritis as core indication

Actemra: stable safety profile over time

10

Event rate / 100 patient-

years over 12-month

periods

012 1324 2536 3748

Serious adverse events 15.7 13.9 15.2 14.4

Serious infections 4.6 3.9 5.2 4.9

Myocardial infarction 0.3 0.2 0.3 0.5

Stroke 0.3 0.1 0.2 0.1

Modified from Genovese, M, et al; Long-Term Safety of Tocilizumab in Rheumatoid Arthritis Clinical Trials.

Arthritis Rheum 2011;63 Suppl 10 :2217

Differentiation versus other medicines in RA

ADACTA in context – what is biologic monotherapy?

RA prevalence

Diagnosed patient

Treated by rheumatologist

Biologic combination therapy –

biologic use with a DMARD

Biologic monotherapy – biologic

use without a DMARD

DMARD(s) + non-DMARD therapy

RA biologics market

(after patients become

DMARD-IR)

Cycling and DMARD combinations (no biologic)

Source: Roche analysis

11

12

Demonstrating differentiation in RA

Monotherapy is a substantial portion of biologic treatment today

12

Monotherapy segment has fewer labeled medicines compared to combination segment

ACTEMRA Enbrel Humira Cimzia Remicade Simponi MabThera Orencia

Indicated in

monotherapy *

Source: Multiple registries on monotherapy (see Prof. Emery presentation); Roche analysis of EU/US labels, *monotherapy label in US only

30%

70%

Biologic

monotherapy

Biologic

combination

Approximately 30% of RA biologics patients are on monotherapy

Demonstrating differentiation in RA

Actemra is already perceived as being different

41 46

16

26 27

66

30 29

0

10

20

30

40

50

60

70

% P

hysic

ian

s

• EU physician perception research conducted Q3 2011

• Physicians were asked to evaluate biologic medicine appropriateness

for use with monotherapy patients “Please rate each biologic on how appropriate you consider it to be for

biologic monotherapy on a scale from 1-10”

% physicians = check top 3 boxes for the product 13

Increasing recognition of Actemra’s strength in monotherapy –

even before ADACTA was published

Actemra monotherapy:

Differentiated efficacy profile vs. other biologic agents

14

Clinical study Treatment arms Key endpoint Key study conclusions

MTX-naïve/free patients

AMBITION (N=503)1 TCZ mono vs MTX ACR20 at Week 24 TCZ mono is superior

to MTX

Patients with active RA despite MTX therapy (MTX-IR)

ACT-RAY (N=553)2 TCZ mono vs TCZ + MTX

DAS28 remission at Week

24

X-ray

Add-on TCZ+MTX

strategy was not

superior to TCZ mono

Patients with inadequate response to DMARDs, including TNF-inhibitors (TNF-IR)

ACT-SURE (N=1681) 3

TCZ 8 mg mono / TCZ

8 mg + MTX or other

DMARDs

Safety at week 24 TCZ mono comparable

to TCZ combo

H2H in Patients with active RA despite MTX therapy (MTX-IR)

ADACTA (N= 325) 4 TCZ mono vs ADA mono Change in DAS28 at Week

24

TCZ mono is superior

to ADA mono

1. Jones G, et al. Ann Rheum Dis 2010; 69:88–96; 2. Dougados M, et al. Ann Rheum Dis 2011; 70(Suppl. 3):73;

3. J. Sibilia, Ann Rheum Dis 2011;70(Suppl3):466. 4. Gabay et al., EULAR12-576

Actemra: Expanding beyond rheumatoid arthritis

15 15

Indication Approved biologic

therapies

Status

• Systemic juvenile

idiopathic arthritis

• None • Actemra received

approval 2011

• Systemic

sclerosis

• None • Phase II – FPI

2012

• Polyarticular

course juvenile

arthritis

• Anti-TNF

inhibitors

• Study read out

2012, filing planned

Additional indications under investigation

Tender

CHERISH

faSScinate

Study name

Rheumatoid arthritis monotherapy in real life

setting: Re-evaluating need and options

Prof. Paul Emery, Director of Musculoskeletal Division, Leeds Institute of Molecular Medicine, University of Leeds, UK

RA patients on biologics do not comply with taking

DMARDs as directed, especially methotrexate

• * Within 90 days before or 90 days after filling biologic prescription

† 92 randomly selected RA patients receiving biologics

• Choquette D et al. Ann

Rheum Dis 2011;

70:197.

45

58

41

54

0

10

20

30

40

50

60

70≥6 months ≥24 months

Pa

tie

nts

(%

)

Time since first biologic prescribed:

DMARD prescription

NOT filled*

MTX prescription

NOT filled*

Anonymous RA patient records (N=6,744) from public and

private drug plans in Canada

17

TNF = tumour necrosis factor;

all registries/studies are anti-TNF focused, other than ORA (abatacept),

AIR (rituximab) and RABBIT (anti-TNFs and anakinra)

ORA5

35%

ARTIS4

30%

NOR-

DMARD1

33%

RABBIT3

34%

AIR5

34%

BSRBR2

32%

Healthcare

insurance claims

database6

30%

CORRONA7

30%

1. Heiberg MS, et al. Arthritis Rheum 2008; 59:234–240. 2. Soliman MM, et al. Ann Rheum Dis 2011; 70:583–589. 3. Listing J, et al. Arthritis Res Ther 2006; 8:R66.

4. Askling J, et al. Ann Rheum Dis 2007; 66:1339–1344. 5. Mariette X, et al. Rheumatology 2011; 50:222–229. 6. Yazici Y, et al. Bull NYU Hosp Jt Dis 2008; 66:77–

85. 7. Lee SJ, et al. J Rheumatol 2009; 36:1611–1617.

One third of RA patients on biologics are on

monotherapy

Data from biologics registries and US claims database

18

Saliot C & van der Heijde D. Ann Rheum Dis 2009;68:1100–1104

n (range)

N (range) 3,463 (24–1,155)

Mean dose of MTX, mg/week (range) 8.8 (4.6–18)

Mean duration of MTX, months (range) 36.5 (27–132)

Adverse event n (range) %

Any adverse event 2,524 (22–475) 72.9%

Permanent discontinuation due to toxicity 315/3,007* 10.5%

Gastrointestinal 1,065 (10–257) 30.8%

Liver 640 (0–122) 18.5%

Skin/hair 309 (0–111) 8.9%

Central nervous system 191 (0–58) 5.5%

Cytopenia 179 (0–27) 5.2%

Lung

- MTX pneumonitis

84 (0–28)

15

2.4%

0.43%

Methotrexate: most common adverse events in

patients with RA

* Total number of patients in studies with data available concerning

permanent discontinuation of MTX

Data from pooled analysis of 21 prospective studies of patients with RA

19

ADACTA: A phase IV study evaluating the reduction in

disease activity during monotherapy treatment with

either Actemra or adalimumab

Prof. Paul Emery, Director of Musculoskeletal Division, Leeds Institute of Molecular Medicine, University of Leeds, UK

C Gabay,1 P Emery,2 R van Vollenhoven,3 A Dikranian,4 R Alten,5

M Klearman,6 D Musselman,6 S Agarwal,6 J Green,7 A Kavanaugh8

1Univ Hospitals of Geneva, Geneva, Switzerland, 2Univ Leeds, Leeds, United Kingdom, 3Karolinska Inst, Stockholm, Sweden, 4San Diego Arthritis Med Clin, San Diego, United

States, 5Univ Berlin, Berlin, Germany, 6Genentech, S San Francisco, United States, 7Roche, Welwyn, United Kingdom, 8Univ California, San Diego, United States

Tocilizumab monotherapy is superior to

adalimumab monotherapy in reducing

disease activity in patients with

rheumatoid arthritis:

24-week data from the phase 4

ADACTA trial

EULAR 2012

June 6-9, 2012; Berlin, Germany

Phase 4, multicentre, randomised, double-blind study of tocilizumab vs adalimumab in RA

22

24 weeks; primary endpoint:

Δ DAS28

TCZ* 8 mg/kg IV Q4 weeks

+ SC Placebo Q2 weeks

1:1 randomisation

Randomised Drug Treatment 8 Weeks Safety Follow-up

Treated

(N = 326) ADA** 40 mg SC Q2 weeks

+ IV Placebo Q4 weeks

Week 16+:

Escapea

aCriteria for escape: <20% improvement from baseline in SJC and TJC at week 16 or after, * TCZ tocilizumab, ** ADA adalimumab

Escape therapy: Weekly SC (ADA/placebo) injections; study medication remained blinded

Superiority trial design

Key inclusion criteria

RA duration of ≥ 6 months

Previous/current MTX* treatment; either MTX-intolerant or continued treatment is considered inappropriate

No prior treatment with a biologic agent

All DMARDs** withdrawn ≥ 2 weeks prior to baseline

Disease Activity Score DAS28 >5.1 at baseline

23

*MTX methotrexate **DMARDs disease-modifying antirheumatic drugs

Study endpoints

Primary endpoint

– Change in DAS28* from baseline to week 24

Key secondary and exploratory endpoints

– Efficacy at week 24: proportions of patients achieving:

DAS28 remission (<2.6) and low disease activity (≤3.2)

ACR20/50/70** responses

ACR/EULAR (Boolean) remission

– Safety: Adverse events and laboratory parameters

Post-hoc analyses

– Clinical Disease Activity Index (CDAI) responses

24

* DAS28 Disease Activity Score (DAS)28, combined index that measures disease activity in patients with RA. It combines information from

28 tender and swollen joints (range0-28), erythrocyte sedimentation rate, and a general health assessment on a visual analog scale. The level

of disease activity is interpreted as low (DAS28≤3.2), moderate (3.2<DAS28≤5.1) or high (DAS28>5.1). DAS28<2.6 corresponds to being in

remission according to the criteria of the American College of Rheumatology.

**ACR20, ACR50, ACR70 represent the percentage of reduction (20%, 50%, 70%) in certain RA symptoms and measures the number of

tender and swollen joints, pain, patient’s and physician’s global assessments and certain laboratory markers.

Patient disposition

25

ADA 40 mg SC Q2 weeks

+ IV placebo Q4 weeks

N = 163*

TCZ 8 mg/kg IV Q4 weeks

+ SC placebo Q2 weeks

N = 163

Randomised patients

N = 326

Completed

125 (77%)

Withdrew

28* (17%)†

Completed

132 (81%)

Withdrew

24 (15%)

Escaped

10 (6%)

Escaped

7 (4%)

*1 ADA patient did not receive study treatment and was not included in the ITT population. †Not including the 2 escape patients that withdrew.

Withdrew from

escape therapy

2 (1%)

Withdrew from

escape therapy

0 (0%)

Demographic and baseline characteristics

26

adalimumab

N = 162

tocilizumab

N = 163

Age, y 53.3 (12.4) 54.4 (13.0)

Female/male, % 82/18 79/21

RA duration, y 6.3 (6.9) 7.3 (8.0)

No. of previous DMARDs 2.0 (1.1) 2.0 (1.1)

Use of oral steroids, % 57 55

Disease activity

DAS28 6.8 (0.9) 6.7 (0.9)

TJC (28 joints) 16.5 (7.0) 15.9 (6.7)

SJC (28 joints) 12.4 (5.4) 11.3 (5.3)

CDAI 43.1 (12.6) 40.8 (12.3)

ESR, mm/h 45.5 (25.4) 50.5 (29.0)

CRP, mg/dl 2.5 (3.9) 2.6 (3.1)

Patient VAS, mm 73.4 (19.4) 71.2 (20.8)

HAQ 1.7 (0.6) 1.6 (0.6)

Data are presented as mean (SD), unless otherwise indicated. List of abbreviations used at the end

Primary endpoint Change in DAS28 from baseline to week 24 (intent-to-treat population)

27

Analysis of variance (model included baseline value plus the stratification factors of region and duration of RA).

1 ADA patient did not receive treatment; 2 TCZ patients had no post-baseline data.

LOCF was used for missing TJC, SJC, ESR, and Patient’s Global VAS. If ESR=0 then ESR=1 was substituted into the

DAS28 calculation to enable a non−missing DAS28

adalimumab 40 mg

SC

+ IV placebo

tocilizumab 8 mg/kg

IV

+ SC placebo

n 162 161

Adjusted mean -1.8 -3.3

Difference -1.5

95% CI for difference -1.8 to -1.1

P-value <0.0001

DAS28 Mean (± SE*) over time

LOCF used for tender and swollen joint counts, ESR and Patient's Global Assessment of Disease Activity VAS

If ESR = 0 then ESR = 1 is substituted into the DAS28 calculation to enable a non-missing DAS28.

8

7

6

5

4

3

2

1

0 Baseline Week

4

Week

8

Week

12

Week

16

Week

20

Week

24

DA

S2

8

ADA 40 mg + placebo (IV) (N = 162) TCZ 8 mg/kg + placebo (SC) (N = 163)

28

*SE standard error

Secondary endpoints Proportions of patients with DAS28 remission/ low disease activity at week 24 (intent-to-treat population)

29

19.8%

10.5%

51.5%

39.9%

0%

10%

20%

30%

40%

50%

60%

DAS28 low disease activity (≤3.2) DAS28 remission (<2.6)

ADA (N = 162) TCZ (N = 163)

*

*

Pa

tie

nts

*P < .0001 (vs ADA); significance was determined using a logistic regression analysis (covariates included treatment,

region, and duration of RA).

LOCF was used for missing TJC, SJC, no imputation was used for ESR and Patient’s Global VAS. Non-responder

imputation was used for missing data. If ESR=0 then ESR=1 was substituted into the DAS28 calculation to enable a

non−missing DAS28.

Secondary endpoints Proportions of patients with ACR20/50/70 response at week 24 (intent-to-treat population)

30

*P < .005 (vs ADA). †P < .0005 (vs ADA).

Significance was determined using a logistic regression analysis (covariates included treatment, region, and duration

of RA).

LOCF was used for missing TJC, SJC. If CRP was missing ESR was substituted.

Non-responder imputation was used for missing data.

*

*

† 49.4%

27.8%

17.9%

65.0%

47.2%

32.5%

0%

10%

20%

30%

40%

50%

60%

70%

ACR 20 ACR 50 ACR70

ADA (N = 162) TCZ (N = 163)

Pa

tie

nts

9.3%

17.2%*

0%

10%

20%

30%

40%

50%

Remission (≥0 to ≤2.8)

31

*P = 0.0389, remission (unadjusted, no control for multiple testing).

†P = 0.0003, remission + low disease activity (unadjusted, no control for multiple testing).

Proportions of patients were compared using CMH analysis stratified by region and duration of RA. Data collected

after withdrawal/initiation of escape therapy was set to missing. LOCF was used for missing data.

Pa

tie

nts

Post-hoc analysis: Clinical Disease Activity Index (CDAI) analysis at week 24 (intent-to-treat population)

ADA (N = 162) TCZ (N = 163)

19.8%

30.7%

Remission + Low Disease Activity

(>0 to <10)

47.9%†

29.0%

Swollen Joint Counts 28 (intent-to-treat population)

0

2

4

6

8

10

12

14

Baseline Wk 4 Wk 8 Wk 12 Wk 16 Wk 20 Wk 24

ADA TCZ

32

Visit

Ab

so

lute

Me

an

S

JC

Co

un

t

(N = 162)

LOCF used for Missing Data

(N = 163)

ADA N = 162;

TCZ N = 163

N = 156;

N = 159

N = 162;

N = 161

N = 162;

N = 161

N = 162;

N = 161

N = 162;

N = 161

N = 162;

N = 161

Safety Adverse events (safety population)

adalimumab

(N=162)

tocilizumab

(N=162)

AEs 443 430

Patients with at least one AE, n (%) 134 (83) 133 (82)

SAEs 21 23

Patients with at least one SAE, n (%) 16 (10) 19 (12)

Infection AEs 106 113

Patients with at least one infection AE, n (%) 68 (42) 77 (48)

Infection SAEs 7 6

Patients with at least one infection SAE, n (%) 5 (3) 5 (3)

Deaths, n (%) 0 (0) 2 (1)

Multiple occurrences of the same AE in 1 individual were counted only once.

33

Safety

Overall, 2 deaths were reported, both in the TCZ arm

– The death of a 49-year old female on study day 2 was considered unrelated to study drug

Cause of death: Illicit drug overdose (marijuana, benzodiazepines, methadone in urine)

– The sudden death of a 56-year old male on study day 93 was considered possibly related to study drug

Comorbidities: Interstitial lung disease, peripheral vascular disease (stent placement), hypertension, overweight (BMI 28), smoking (for 30 years)

Cause of death: Unknown (no autopsy was performed)

34

35

0

1

2

3

4

Baseline Wk 8 Wk 16 Wk 24

ADA TCZ

Visit

mm

ol/

L

mg

/dL

160

120

80

40

LDL-cholesterol by visit (safety population)

ADA n = 144 TCZ n = 146

ADA n = 127 TCZ n = 128

ADA n = 137 TCZ n = 138

ADA n = 138 TCZ n = 143

Patients on LLA at baseline & no change in LLA post BL = 20 patients in each arm

Patients started on LLA after week 4 = 6 ADA patients; 10 TCZ patients

LD

L-c

ho

les

tero

l

ALT levels by CTC* grade (worst value) (safety population)

72.2%

24.7%

1.9% 1.2%

62.3%

30.9%

5.6% 1.2%

0%

10%

20%

30%

40%

50%

60%

70%

80%

Normal >ULN-2.5xULN >2.5x-5xULN >5x-20xULN

ADA (N=162) TCZ (N=162)

Pa

tie

nts

, %

*CTC Common Toxicity Criteria: Grade 1, >ULN to 2.5xULN; Grade 2, >2.5xULN to 5xULN; Grade 3, >5xULN to 20xULN.

ALT levels

36

Conclusions

Efficacy

– Tocilizumab (TCZ) monotherapy was superior to adalimumab (ADA) monotherapy in reducing signs and symptoms of RA

Safety

– The overall adverse event profile was comparable between the two treatment arms

– The safety observed in the TCZ arm of this study is consistent with the known safety profile of TCZ and no new or unexpected adverse events were observed

37

Investigators (82 centres from 15 countries)

38

Australia: Nash Youssef

Belgium: Malaise Margaux

Brazil: Radominski Ximenes Zerbini

Czech Republic: Pavelka

Finland: Hannonen Leirisalo-Repo

Germany: Alten Braun Fiehn Gauler Heilig

Rubbert-Roth Specker Tony Wassenberg von Hinuber

Greece: Aslanidis Boumpas Sfikakis

Mexico: Elizonda Irazoque Zazueta

Portugal: Bernardes Canas Da Silva Fonseca

Spain: Alvaro Gracia Blanco Juan Navarro-Sarabia Sanmarti

Sweden: Baecklund van Vollenhoven

Switzerland: Dudler Gabay Hasler Kyburz Muller

Turkey: Bilgen Hamuryudan Karaaslan Terzioglu

UK: Emery Hakim Isaacs Sheeran Thompson

USA: Borofsky Bushan Chindalore Churchill Curtis

Dikranian Ettlinger Forstot Fung Gladstein

Halter Hensarling Huffstutter Hsu Jerdan

Kenney King Ii Kivitz Klein Kohen

Lawson Lee Logan Lue Payne

Rizzo Samuels Sayers Scoville Sherrer

Singhal Trapp Waller Wolfe

Disclosures

The authors disclose the following financial relationships:

– Cem Gabay Consultant for: Roche2,4, Abbott2,4, Merck2,4, UCB 2,4, Pfizer, 2,4 BMS2,4, Merckserono 2,4, Novartis2,4, Amgen 2,4

– Paul Emery: Merck4, Abbott4, Pfizer4, Roche4, UCB 4, BMS 4

– Ronald van Vollenhoven: Abbott1,4, GSK1,4, MSD1,4, Pfizer 1,4, Roche1,4, UCB1 ,4

– Alten Dikranian: Genentech 2, UCB2, Abbott2, BMS2

– Rieke Alten: BMS1,2,4, Novartis1,2,4, Pfizer1,2,4, Roche1,2,4, UCB1,2,4, Abbott2,4

– Micki Klearman: Genentech3

– David Musselman: Genentech3

– Sunil Agarwal: Genentech3

– Jennifer Green: Roche3

– Arthur Kavanaugh: Roche1, Amgen1, Abbott1, BMS1, Janssen1, UCB1, Pfizer1

1Research Grants; 2 Speakers Bureau; 3 Employment; 4 Consulting Fees

39

Actemra/RoActemra summary

Dr. Karsten Jung, Global Product Strategy Head for Immunology & Ophthalmology, Roche

Actemra clearly demonstrates differentiation in RA

Timeline Indication Milestone

2012

RA DMARD IR 1st line biologic

label in US Expect FDA approval

RA DMARD IR monotherapy Ph III ADACTA H2H vs. Humira

Application to EMA for label update

RA, moderate to severe;

subcutaneous application

Ph III SUMMACTA

Ph III BREVACTA

Application to FDA and EMA

RA, early moderate to severe Ph III FUNCTION

Polyarticular-course juvenile

idiopathic arthritis

Ph III CHERISH

Application to FDA and EMA

2013 Systemic sclerosis Ph II readout

41

• Actemra has now a strong foundation of clinical data in RA

• ADACTA provides solid H2H data to inform treatment decisions in RA monotherapy

• Upcoming clinical and regulatory newsflow will advance the franchise

Questions & Answers

42

List of abbreviations

43

ACR20/50/70 Percentage of reduction (20%, 50%, 70%) in certain

RA symptoms; also measures the number of tender

and swollen joints, pain, patient’s and physician’s

global assessments and certain laboratory markers

ADA Adalimumab

AE Adverse event

ALT Alanin transaminase; liver enzyme used for

diagnostic evaluation of hepatic injury

Anti-TNF inh Anti-tumor necrosis factor inhibitor

CDAI Clinical disease activity index; score for

routine assessment of disease activity in

rheumatoid arthritis patients

CTC Common toxicity criteria

DAS28 Disease activity Score 28; index combines information

from 28 tender and swollen joints (range0-28),

erythrocyte sedimentation rate, and a general health

assessment on a visual analog scale. The level of dis-

ease activity is interpreted as low (DAS28≤3.2), mode-

rate (3.2<DAS28≤5.1) or high (DAS28>5.1).

DAS28<2.6 corresponds to being in remission accor-

ding to the criteria of the American College of

Rheumatology

DMARD Disease-modifying antirheumatic drug

DMARD-naive No previous administration of DMARD

ESR Erythrocyte sedimentation rate

FPI First-patient-in

H2H Head-to-head

HAQ Health assessment questionnaire

IR Inadequate responder

ITT Intent-to-treat

IV Intravenous

LDL Low-density lipoprotein

LLA Lipid-lowering agent

LOCF Last observation carried forward: analysis method in

rheumatoid arthritis

MTX Methotrexate

MTX-naive No previous administration of methotrexate

RA Rheumatoid arthritis

SAE Serious adverse event

SC Subcutaneous

SE Standard error

SJC 28 Swollen joint count 28

TJC 28 Tender joint count 28

TCZ Tocilizumab

ULN Upper limit of normal

VAS Visual analogue scale, measurement instrument for

subjective characteristics

44

We Innovate Healthcare

Appendix

45

Actemra clinical safety

46

Event rate / 100 patient-years

over 12-month periods 012 1324 2536 3748

Serious adverse events 15.7 13.9 15.2 14.4

Serious infections 4.6 3.9 5.2 4.9

Myocardial infarction 0.3 0.2 0.3 0.5

Stroke 0.3 0.1 0.2 0.1

4,009 ACTEMRA patients in clinical trials, median duration was 3.6 years, total observation time was 12,293 patient-years*

Rates of SAEs, serious infections and cardiovascular events remained stable over time

Actemra Epidemiology

data for TNF-

antagonists 6 month pooled controlled data LTE studies

Data cut Feb 2010 Placebo/DMARD All TCZ PY exposure 628 1132 12293 3800 a

Rate of deaths

(95% CI)

Number of events

0.80

(0.26, 1.86)

5

0.44

(0.14, 1.03)

5

0.56

(0.44, 0.71)

69

0.61

(0.38, 0.91) a

23

*Modified from Genovese, M, et al; Long-Term Safety of Tocilizumab in Rheumatoid Arthritis Clinical Trials. Arthritis Rheum 2011;63 Suppl 10 :2217

Mortality rates of patients treated with Actemra are no greater than those observed with

TNF antagonists

a Unadjusted mortality rate for RA patients treated with TNF antagonists; calculated based on meta-analysis data of 17 trials; 4097 patients

Leombruno et al. Ann Rheum Dis 2009;68:1136-1145

Rate of Deaths per 100 Patient Years