1. 1 Natalizumab Safety Gordon Francis, MD Senior Vice President, Clinical Development

2. 2 Natalizumab Safety  Exposure  General safety overview  Infections  Other safety events of interest based on MOA  Post-marketing safety during 2nd MS launch in June 2006

3. 3 Safety Population Clinical Development Program MS Experience Placebo n=1135 Natalizumab n=1617 n=2321 3321 patient-years CD Experience Placebo n=506 Natalizumab n=1182 n=1563 1338 patient-years Cumulative MS and CD 3884 patients 4659 patient-years

4. 4 CD Studies Used in Safety Analysis  Short-term placebo-controlled studies of active CD –Generally 3 months of dosing –>80% of patients were in CD301 and CD307 431 placebo patients and 983 natalizumab patients –The remainder were from the Phase 2 Studies  Short and long-term dosing –1563 patients exposed to natalizumab for up to 3 years –Includes placebo-controlled and open-label data

5. 5 Long-Term Natalizumab Exposure During CD351 During CD351 and Feeder Studies Number of patients1100 Mean number monthly infusions (SD) 11.0 (8.6) 14.8 (10.9) Median812 Range1-241-39

6. 6 General Safety

7. 7 Overview of Adverse Events in CD Short-Term, Placebo-Controlled Studies of Active CD Placebo N=506 % Natalizumab N=1182 % AEs85.687.4 Infections36.240.4 SAEs14.014.9 Non-Crohn’s SAEs 7.810.1 Serious infections 2.4 Serious systemic hypersensitivity0 0.1 Malignancy 0.2 0.6

8. 8 SAEs by Organ System Excluding Crohn’s Disease Placebo-Controlled Studies of Active CD SAE Placebo N=506 % Natalizumab N=1182 % Overall7.810.1 Gastrointestinal2.43.9 Infection2.4 Musculoskeletal0.6 Nervous system0.20.5 General0.80.6 Cardiac00.4 Immune system0.20.4 Hepatobiliary00.4

9. 9 Placebo N=506 % Natalizumab N=1182 % Total patients discontinued11.39.1 Crohn’s disease7.93.7 Urticaria0.20.8 Abdominal pain0.00.4 Dyspnea0.00.3 Flushing0.00.3 Nausea0.20.3 Pruritus0.20.3 Discontinuations Due to AEs Short-Term Placebo-Controlled Studies of Active CD

10. 10 Deaths in CD Program Placebo-Controlled Experience Age/GenderCause of DeathInfusion # Mo. Since Infusion Medications/ Relevant History 42/MAsphyxiation1—None Open-Label Experience Age/GenderCause of DeathInfusion# Mo. Since Infusion Medications/ Relevant History 49/FPeritonitis, renal failure 31Post-op complications 60/MPML86AZA, lymphopenia 75/MPulmonary aspergillosis 103Prednisolone, GI bleed, hospitalization 67/MMI, shock282Hypertension 69/MPCP343Cirrhosis, ARF, sepsis

11. 11 Infections  General overview of infections –Serious –Risk of infection over time –Infection with concomitant medication use –Viral reactivation Herpes family –Opportunistic infections, including PML

12. 12 Serious Infections Short-Term Placebo-Controlled Studies of Active CD Placebo N=506 % Natalizumab N=1182 % Overall2.4 Perianal abscess0.6 Abdominal abscess0.20.3 Gastroenteritis0.2 Meningitis viral00.2 Urinary tract infection00.2 Individual events only listed for serious infections occurring in ≥ 2 natalizumab-treated patients

13. 13 Incidence and Rate of Infection Over Time Short- and Long-Term Dosing Time interval (months) 0-6 n=1563 7-12 n=681 13-18 n=529 19-24 n=427 25-30 n=294 Incidence (%)48.948.042.038.228.9 Exposure (person-years) 739307241198122 Rate of infection per person-year 1.8 1.41.20.9 Most common infections – nasopharyngitis, influenza, URTI, sinusitis, viral infection

14. 14 Rate of Serious Infection Over Time Short- and Long-Term Dosing Serious Infection Rate (95% CI) 0-66-1212-1818-2424-31 0.00 n=1563 Months 0.02 0.04 0.06 0.08 0.10 n=681n=427n=294n=509

15. 15 Incidence of Infection with Concomitant Medication Placebo-Controlled Studies in Active CD Patients With an Infection (%) 0 10 20 30 40 50 n=340 39 n=264 41 n=373 41 n=205 40 Natalizumab alone Natalizumab + immunosuppressants and steroids Natalizumab + steroids Natalizumab + immunosuppressants

16. 16 Serious Infections with Concomitant Medication Use Short-Term, Placebo-Controlled Studies of Active CD n Placebo/ Natalizumab Placebo % Natalizumab % Monotherapy154/3731.31.9 + Immunosuppressants89/2051.13.9 + Steroids154/3401.32.6 + IS + Steroids109/2646.41.9 IS=Immunosuppressants

17. 17 Incidence of Herpetic Infections Short-term Placebo-Controlled Studies of Active CD Placebo n=506 Natalizumab n=1182 Total patients with herpes infections6(1.2)20(1.7) Herpes simplex4 (0.8) 15(1.3) Herpes zoster1 (0.2) 4(0.3) Herpes viral infection NOS01(<0.1) CMV infection (colitis)01(<0.1) Number (%) of Patients

18. 18 Summary of Serious Herpes Infections Short- and Long-term CD Population Patients Age/Gender Infection TypeInfusion# Meds/ Relevant Hx Outcome 36/FHSV Vaginitis0Developed pre-treatment Acyclovir, recovered 12/MHerpes zoster1Developed pre-treatment Acyclovir, recovered 33/FCMV colitis2 AZARecovered 33/MCMV infection1 6MPRecovered 30/FVaricella pneumonia8Son with chicken pox Acyclovir, recovered 25/MHerpes zoster14Prednisolone, AZAAcyclovir, recovered

19. 19 Opportunistic and Atypical Infections

20. 20 Opportunistic Infections Cumulative CD Experience Patient Age/Gender Infection TypeInfusion# Meds/ Medical Hx Outcome 62/FBurkholderia cepacia pneumonia 3Tobacco use, CHF, steroids Recovered 65/FMAC infection8Steroids / S.aureus pneumonia Recovered 60/MPML 8 AZA / lymphopeniaDeath 75/MPulmonary aspergillosis 10 Steroids / GI bleed, prolonged hospitalization Death 69/MPCP34 Cirrhosis, ARF, sepsisDeath

21. 21 Serious Atypical Infections Cumulative CD Experience Age/ Gender Infection Type Infusion # Months since last Infusion Meds/ Medical Hx Outcome 33/F CMV Colitis 21AzathioprineRecovered 33/MCMV Infection166MPRecovered 48/FCandida Sepsis 14Steroids, azathioprine/multi-organ failure Recovered 30/FVaricella pneumonia 8—Son with chicken-poxRecovered 32/MCavitating pneumonia 13—Azathioprine, corticosteroids Recovered 20/MTb Peritonitis225Caseating granuloma; AFB, PCR, culture neg Recovered

22. 22 Rate of Non-PML Opportunistic Infections in CD Rate/1000 pt-yrs Unique opportunistic infections0.6* Opportunistic infections seen in CD patients on conventional therapies or TNF-  Inhibitors 0.3-0.5** Opportunistic infections seen in CD patients not on therapy 0.06-0.2** **Health Benchmarks International database * Based on 1771 patient-years

23. 23 PML in Natalizumab-Treated Patients  A total of 3 confirmed cases of PML –2 MS patients after 29 and 37 infusions Both received IFNβ concurrently –1 CD patient after 8 infusions over 18 months Long-standing Grade 3 lymphopenia  2 cases fatal, 1 survived following development of immune reconstitution inflammatory syndrome (IRIS)

24. 24 Overview of PML  Rare, progressive infection of the CNS –Often fatal within 6 months of diagnosis  JC virus, a latent human polyomavirus, produces lytic infection of oligodendroglia as the final step in multi-step process  Primarily affects immunocompromised individuals –Hematologic malignancies –AIDS patients –Organ transplantation  Therapy –No effective therapy –Improvement seen with IRIS (immune reconstitution inflammatory syndrome) seen in HIV-related PML with HAART or in transplant- related PML with reduction of immune suppressants

25. 25 PML in Crohn’s Patient  60-year-old male with Crohn’s disease for 28 years –History of treatment with azathioprine, corticosteroids, infliximab –Longstanding lymphopenia (<500) –Natalizumab dosing over 18 months 3 doses of natalizumab + azathioprine (CD301) 9 doses of placebo + azathioprine (CD303) 5 doses of natalizumab ending in June 2003 (CD351) –Pathological diagnosis of astrocytoma in July 2003 –Progressive worsening leading to death in December 2003 –Pathology reassessment in March 2005: diagnosis of PML

26. 26 Independent PML Safety Evaluation  All dosing suspended 28 February 2005  Objectives of safety evaluation –Determine if additional patients had undiagnosed PML or other atypical infections  3819 MS, CD and RA patients eligible for the evaluation  Vital status confirmed in >99%  90% of natalizumab-treated patients participated –Neurological examination, brain MRI, JCV in plasma and CSF  No new cases of PML

27. 27 Options for PML Monitoring  JC viral DNA in plasma –0/10 patients with JCV detected in blood had MRI or clinical findings suggestive of PML –Only 1/3 PML cases had JCV in blood pre-symptomatically –Sensitivity and predictive value of JCV testing is low  MRI brain scans –Sensitive but not specific –Large numbers of scans needed to pre-symptomatically detect rare event –Helpful diagnostically but not as a practical PML screening test  Clinical vigilance –Patient and healthcare provider education –Monthly pre-infusion questions

28. 28 Other Safety Events of Interest Infusion Reactions Malignancy Post-Marketing Safety

29. 29 Hypersensitivity and Infusion Reactions – CD  3.5% of patients in CD301 and 307 experienced hypersensitivity reactions* frequently associated with antibody formation, usually with 2nd or 3rd infusion  Serious systemic hypersensitivity reactions –In short-term studies 0.1% natalizumab –In short- and long-term dosing population 0.1% natalizumab * Hypersensitivity, urticaria, anaphylaxis, pruritus.

30. 30 Malignancy

31. 31 Event Placebo N=506 Natalizumab N=1182 Overall n (%)1 (0.2)7 (0.6) n Adenocarcinoma of lung02 Bladder cancer01 Breast cancer01 Breast cancer invasive01 Colon cancer01 Malignant melanoma01 Uterine cancer1 0 Rate per 100 person-years: 0.60 for placebo; 1.6 for natalizumab Malignancy Short-Term Placebo-Controlled Studies of Active CD

32. 32 Summary of Safety in Clinical Development  Common adverse events were comparable between treatment groups and generally mild  Proportions of patients experiencing SAEs comparable  Infections slightly more common with natalizumab  Low rate of HS reactions  Risk of opportunistic infection exists  Possible signal for malignancy that requires post-marketing surveillance measures

33. 33 Post-Marketing Safety

34. 34 Post-Marketing Safety Data  ~11,500 MS patients world-wide have received natalizumab commercially between June 2006 and May 2007  Serious infections –Reporting rate = 1.2% (comparable to clinical trials 2.6-3.2%) –Urinary and pulmonary infections most common (0.3%) –4 cases of atypical herpes infections Encephalitis, meningitis during 1st launch Multidermatomal zoster and HSV1 esophagitis, considered as OIs during 2nd launch Serious herpes infections in label  Serious hepatic dysfunction –4 cases with marked elevation of LFTs, including bilirubin in 3 of 4 cases –Potential confounding factors – evaluation continuing –No signal seen in clinical development (adverse events or LFT abnormalities)

35. 35 Post-Marketing Safety Data  Malignancy –Reporting rate = 6.1/1000 patient years (comparable to clinical trials (3.8-7.3) and SEER database (5.0))  Hypersensitivity reactions more common in those previously dosed than those naïve to therapy –Label change being implemented to highlight this for practitioners  2 cases of atypical herpes infection reported –Multi-dermatomal herpes zoster, HSV esophagitis  No new confirmed case of PML  No change in safety profile of natalizumab in post- marketing setting compared with clinical trials

36. 36 Favorable Benefit-Risk Profile  Efficacy for inducing and maintaining response and remission –Benefit consistent in broad range of subgroups  Safety profile comparable to that seen in MS indication apart from potential increased risk of non-PML opportunistic infections  RiskMAP in place for MS will be adapted for CD to further evaluate long-term risk