XELJANZ Tofacitinib in Rheumatoid Arthritis EUGENE LIM MBBS

XELJANZ (Tofacitinib) in Rheumatoid Arthritis EUGENE LIM MBBS, MMed (Int Med), FAMS (Rheum) FACR, Rh. MSUS, RMSK

The JAK-STAT Pathway

JAK Pathway Signalling Cytokines 1. Cytokines bind to the cell-surface receptors 2. JAKs are activated and JAK phosphorylated with ATP 3. STATs bind at the receptor P P JAK STAT and are phosphorylated by activated JAKs P P 4. STATs translocate to the nucleus to alter gene transcription Gene Transcription/Cytokine Production JAK, Janus kinase inhibitors; ATP, adenosine triphosphate; P, phosphate; STAT, signal transducer and activator of transcription O’Shea JJ et al. N Engl J Med. 2013; 368: 161 -170.

JAK Inhibition Modulates Cytokine Signaling From Inside the Cell Cytokines 1. Cytokines bind to cell-surface JAK inhibitor receptors 1 2. JAK inhibitors inhibit JAK phosphorylation and activation of JAKs 2, 3 3. JAKs cannot phosphorylate the JAK STAT cytokine receptors 2, 3 4. STATs cannot dock, and are not phosphorylated or activated 2, 3 5. Gene transcription and cytokine production is partially inhibited 2, 4 Reduced Gene Product/Cytokine Production JAK, Janus kinase inhibitors; STAT, signal transducer and activator of transcription 1. O’Shea JJ et al. N Engl J Med. 2013; 368: 161 -170; 2. XELJANZ [prescribing information]. New York, NY: Pfizer Inc; 2016; 3. Jiang JK et al. J Med Chem. 2008; 51: 8012 -8018; 4. Ghoreschi K et al. J Immunol. 2011; 186: 4234 -4243.

JAK Inhibition by Tofacitinib

Tofacitinib, as a Pan-JAK Inhibitor, Modulates the Signaling of Multiple JAK-Dependent Cytokine Families IC 50, n. M, [ATP]=1 m. M 4 JAK 1/JAK 2 JAK 1/JAK 3 JAK 2/JAK 2 406 56 1377 γ-chain cytokines IL-2, IL-4, IL-7, IL-9, IL-15, IL-21 IL-10 a IL-22 IFN- / IL-6 IL-11 IFN-γ IL-12 IL-23 EPO TPO GM-CSF TYKb JAK 2 b JAK 1 JAK 3 TYK 2 JAK 1 JAK 2 TYK 2 JAK, Janus kinase inhibitors; TYK 2, Tyrosine kinase 2; IL, interleukin ; IFN, interferon; EPO, erythropoietin; GM-CSF, Granulocyte-macrophage colony-stimulating factor; TPO, Thrombopoietin; IC 50, half maximal inhibitory concentration; n. M, nanomolar; m. M, millimolar a. IL-10/IL-22 may have pro- or anti-inflammatory activities depending on the cellular environment and/or disease state. 3 II cytokine receptors such as those for gp 130 subunit sharing receptors for IL-6 and IL-11 as well as IL-10, IL-19, IL-20, and IL-22, mainly signal through JAK 1, but also associate with JAK 2 and TYK 2. 2, 3 1. O’Sullivan LA et al. Mol Immunol. 2007; 44: 2497 -2506; 2. Ghoreschi K et al. Immunol Rev. 2009; 228: 273287; 3. Sanjabi S et al. Curr Opin Pharmacol. 2009; 9: 447 -453; 4. XELJANZ Malaysia PI, XELJANZ-0117 b. Type

Most JAK-Dependent Cytokines Use JAK 1 a JAK 1 Inhibitor Has a Broad Cytokine Inhibition Profile Receptors sharing γ-chain 1 Type I IFNR 1 IL-10 R family 1 Receptors sharing gp 130 subunit 1, 2* Type II IFNR 1 IL-12 R family sharing p 40 subunit 1 Hormone receptors 1 IL-3 R family 1 IL-2, IL-4, IL-7, IL-9, IL-15, IL-21 IFNα/β/κ/ω /ε IL-10†, IL-20, IL-22 IL-6, IL-11, IL-27, G-CSF IFN-γ IL-12, IL-23 EPO, TPO, GH, PRL IL-3, IL -5 TYK 2* JAK 3 JAK 1 TYK 2 JAK 1 JAK 2* JAK 2 JAK 1 TYK 2 JAK 1 inhibitor + + - - JAK 2 inhibitor - - + + JAK 3 inhibitor + - - - TYK 2 inhibitor - + + - All of these cytokines ultimately signal through the 6 signal transducer and activator of transcription (STAT) proteins JAK, Janus kinase inhibitors; TYK 2, Tyrosine kinase 2; IL, interleukin ; IFNR, interferon; ; GM-CSF, Granulocytemacrophage colony-stimulating factor; TPO, Thrombopoietin; GH, growth hormone; PRL, prolactin *Type II cytokine receptors such as those for IL-10, IL-19, IL-20, and IL-22 as well as gp 130 subunit sharing receptors for IL-6 and IL-11 mainly signal through JAK 1, but also associate with JAK 2 and TYK 2. 2 †IL-10/IL-22 may have pro- or anti-inflammatory activities depending on the cellular environment and/or disease state. 4 1. O’Sullivan LA, et al. Mol Immunol. 2007; 44: 2497 -506; 2. Ghoreschi K, et al. Immunol Rev. 2009; 228: 273 -287; 3. Ghoreschi K et al. Nat Immunol. 2009; 10: 356 -360.

Overview of the ORAL trials

Tofacitinib has been Evaluated in One of the Most Extensive RA Clinical Development Programs to Date 1 ORAL Start 2 ORAL Solo 3 ORAL Sync 4 ORAL Standard 5 ORAL Scan 6 ORAL Step 7 Patient Type MTX naïve DMARD-IR MTX-IR TNFi-IR Distinguishing Feature Radiographic data/ monotherapy Monotherapy Background DMARDsa Active controlb (adalimumab) Radiographic data TNFi-IR Study Duration, mo 24 6 12 12 24 6 Treatment XELJANZ 5 or 10 mg bid + nonbiologic DMARD(s) XELJANZ 5 or 10 mg bid + MTX Patients Enrolled 952 c 610 795 717 797 399 • ACR 20 • HAQ-DI • DAS 28 -4(ESR) <2. 6 • ACR 20 • HAQ-DI • DAS 284(ESR) <2. 6 • m. TSS • ACR 20 • HAQ-DI • DAS 284(ESR) <2. 6 Kremer J et al. Ann Intern Med. 2013 van Vollenhoven RF et al. N Engl J Med. 2012 van der Heijde D et al. Arthritis Rheum. 2013 Primary Endpoints Publication • ACR 70 • m. TSS Lee EB et al. N Engl J Med. 2014 Fleischmann R et al. N Engl J Med. 2012 Burmester GR et al. Lancet. 2013 XELJANZ is not indicated in MTX-naïve patients. The recommended dose of XELJANZ is 5 mg twice daily. ACR=American College of Rheumatology; DMARD-IR=disease-modifying antirheumatic drug-inadequate responder; MTX-IR=methotrexate-inadequate responder; TNFi. IR=tumor necrosis factor inhibitor inadequate responder. a. Nonbiologic DMARDs. b. ORAL Standard was not designed to provide head-to-head comparative efficacy data and should not be interpreted as evidence of superiority or noninferiority to adalimumab. . c 958 patients were randomized. Two in the 10 -mg group were never treated; 956 received at least 1 dose of medication. The year 1 prespecified interim analysis included 952 patients (371 XELJANZ 5 mg bid; 395 XELJANZ 10 mg bid; 186 MTX). Final data on 2 additional patients in each XELJANZ group was available for final analyses. 1. XELJANZ [prescribing information]. New York, NY: Pfizer Inc; 2016. 2. Lee EB et al. N Engl J Med. 2014; 370(25): 2377 -2386. 3. Fleischmann R et al. N Engl J Med. 2012; 367(6): 495 -507. 4. Kremer J et al. Ann Intern Med. 2013; 159(4): 253 -261. 5. van Vollenhoven RF et al. N Engl J Med. 2012; 367(6): 508 -519. 6. van der Heijde D et al. Arthritis Rheum. 2013; 65(3): 559 -570. 7. Burmester GR et al. Lancet. 2013; 381(9865): 451 -460.

Proportion (%) of ACR 20 responders across studies (FAS, NRI) *** *** *** ** * N= 154 309 106 ORAL Scan 1 (6 mos) 196 199 ORAL Standard 2, 3 (6 mos) MTX-IR 157 309 120 ORAL Sync 4 (6 mos) 242 ORAL Solo 5 (3 mos) DMARD-IR 131 ORAL Step 6 (3 mos) TNFi-IR *P≤ 0. 05; **P<0. 001; ***P<0. 0001 vs. placebo/MTX. Tofacitinib 5 mg BID is the approved dose. ORAL standard study was not designed for non-inferiority/superiority comparisons between tofacitinib and adalimumab. 1. van der Heijde D, et al. Arthritis Rheum 2013; 65: 559 -570. 2. Pfizer Inc. FDA briefing document: Tofacitinib for the treatment of rheumatoid arthritis (NDA 203214; 9 May 2012). 2012. Available at: http: //www. fda. gov/downloads/Advisory. Committees/Committees. Meeting. Materials/ Drugs/Arthritis. Advisory. Committee/UCM 302960. pdf. Accessed 13 October 2015. 3. van Vollenhoven RF, et al. N Engl J Med 2012; 367: 508 -519. 4. Kremer J, et al. Ann Intern Med 2013; 159: 253 -261. 5. Fleischmann R, et al. N Engl J Med 2012; 367: 495 -507. 6. Burmester GR, et al. Lancet 2013; 381: 451 -460. For education purpose only. Not for distribution. 133

Proportion (%) of ACR 50 responders across studies (FAS, NRI) *** *** N= 154 *** ** 309 106 ORAL Scan 1 (6 mos) 196 199 ORAL Standard 2, 3 (6 mos) MTX-IR 157 *** 309 120 ORAL Sync 4 (6 mos) 242 ORAL Solo 5 (3 mos) DMARD-IR *** 131 ORAL Step 6 (3 mos) TNFi-IR **P<0. 001; *** P<0. 0001 vs. placebo/MTX. Tofacitinib 5 mg BID is the approved dose. ORAL standard study was not designed for non-inferiority/superiority comparisons between tofacitinib and adalimumab. 1. van der Heijde D, et al. Arthritis Rheum 2013; 65: 559 -570. 2. Pfizer Inc. FDA briefing document: Tofacitinib for the treatment of rheumatoid arthritis (NDA 203214; 9 May 2012). 2012. Available at: http: //www. fda. gov/downloads/Advisory. Committees/Committees. Meeting. Materials/ Drugs/Arthritis. Advisory. Committee/UCM 302960. pdf. Accessed 13 October 2015. 3. van Vollenhoven RF, et al. N Engl J Med 2012; 367: 508 -519. 4. Kremer J, et al. Ann Intern Med 2013; 159: 253 -261. 5. Fleischmann R, et al. N Engl J Med 2012; 367: 495 -507. 6. Burmester GR, et al. Lancet 2013; 381: 451 -460. For education purpose only. Not for distribution. 133

Proportion (%) of ACR 70 responders across studies (FAS, NRI) *** *** * N= 154 309 106 ORAL Scan 1 (6 mos) 196 199 ORAL Standard 2, 3 (6 mos) MTX-IR * *** 157 309 120 ORAL Sync 4 (6 mos) 242 ORAL Solo 5 (3 mos) DMARD-IR ** 131 ORAL Step 6 (3 mos) TNFi-IR *P≤ 0. 05; **P<0. 001 ; *** P<0. 0001 vs placebo/MTX. Tofacitinib 5 mg BID is the approved dose. ORAL standard study was not designed for non-inferiority/superiority comparisons between tofacitinib and adalimumab. 1. van der Heijde D, et al. Arthritis Rheum 2013; 65: 559 -570. 2. Pfizer Inc. FDA briefing document: Tofacitinib for the treatment of rheumatoid arthritis (NDA 203214; 9 May 2012). 2012. Available at: http: //www. fda. gov/downloads/Advisory. Committees/Committees. Meeting. Materials/ Drugs/Arthritis. Advisory. Committee/UCM 302960. pdf. Accessed 13 October 2015. 3. van Vollenhoven RF, et al. N Engl J Med 2012; 367: 508 -519. 4. Kremer J, et al. Ann Intern Med 2013; 159: 253 -261. 5. Fleischmann R, et al. N Engl J Med 2012; 367: 495 -507. 6. Burmester GR, et al. Lancet 2013; 381: 451 -460. For education purpose only. Not for distribution. 133

ACR 50 response as early as 2 weeks¥ ORAL Step 3 ORAL Sync 1, 2 Tofacitinib 5 mg BID Placebo *p≤ 0. 001 vs baseline; †p≤ 0. 01 vs placebo *p≤ 0. 05; †p<0. 0001 vs placebo ¥ ACR 50 was one of the secondary efficacy endpoints in the sutdy. ACR=American College of Rheumatology; BID=twice daily. Tofacitinib 5 mg BID is the approved dose. 1. Kremer J, et al. Ann Intern Med 2013; 159: 253 -261. 2. Kremer J, et al. Supplement to: Ann Intern Med 2013; 159: 253 -261. 3. Burmester GR, et al. Lancet 2013; 381: 451 -460. For education purpose only. Not for distribution.

Proportion of Patients with DAS-28 -Defined Remission∫ (FAS, NRI) * † * * MTX-IR * DMARD-IR TNFi-IR p≤ 0. 05; *** p<0. 0001 vs. Placebo (unadjusted). † Statistical significance could not be declared in the ORAL Scan study due to the step down procedure. FAS: Full analysis Set; NRI: Nonresponder Imputation. ; SE: Standard Error; ORAL standard study was not designed for non-inferiority/superiority comparisons between tofacitinib and adalimumab. Tofacitinib 5 mg BID is the approved dose. 1. Burmester G, et al. Lancet. 2013; 381: 451– 460; 2. Fleischmann R, et al. New Engl J Med. 2012; 367: 495 -507; 3. Kremer J et al. Ann Intern Med. 2013; 159(4): 253 -61; 4. Van der Heijde D, et al. Arthritis Rheum 2013; 65(3): 559 -570; 5. van Vollenhoven RF, et al. New Engl J Med. 2012; 367: 508 -19. ; 6. Lee EB, et al Arthritis Rheum 2012; 64(10 [supplement]): S 1049, Abst 2486. For education purpose only. Not for distribution. 14

Mean Change from Baseline in HAQ-DI Scores (FAS, Mixed-effect longitudinal linear model) MTX-IR ORAL Scan (3 mos) DMARD-IR ORAL Standard (3 mos) ORAL Sync (3 mos) ORAL Solo (3 mos) TNFi-IR ORAL Step (3 mos) MCID † *** ** ** p≤ 0. 001; *** p<0. 0001 vs. Placebo. , †Statistical significance could not be declared in the ORAL Scan study due to the step down procedure FAS: Full Analysis Set; MCID: Minimum Clinically Important Difference; HAQ: Health Assessment Questionnaire’; DI: Disability Index; Mos: months; ORAL standard study was not designed for non-inferiority/superiority comparisons between tofacitinib and adalimumab. 1. Burmester G, et al. Lancet. 2013; 381: 451– 460; 2. Fleischmann R, et al. New Engl J Med. 2012; 367: 495 -507; 3. Kremer J et al. Ann Intern Med. 2013; 159(4): 253 -61; 4. Van der Heijde D, et al. Arthritis Rheum 2013; 65(3): 559 -570; 5. van Vollenhoven RF, et al. New Engl J Med. 2012; 367: 508 -19. ; 6. Lee EB, et al Arthritis Rheum 2012; 64(10[supplement]): S 1049, Abst 2486 For education purpose only. Not for distribution.

ORAL Strategy

Efficacy and safety of tofacitinib monotherapy, tofacitinib with methotrexate, and adalimumab with methotrexate in patients with rheumatoid arthritis (ORAL Strategy): a phase 3 b/4, double-blind, head-to-head, randomised controlled trial Roy Fleischmann, Eduardo Mysler, Stephen Hall, Alan J Kivitz, Robert J Moots, Zhen Luo, Ryan De. Masi, Koshika Soma, Richard Zhang, Liza Takiya, Svitlana Tatulych, Christopher Mojcik, Sriram Krishnaswami, Sujatha Menon, Josef S Smolen, on behalf of the ORAL Strategy investigators Fleischmann R et al. Lancet 2017; 390: 457 -68; Published online June 16, 2017; http: //dx. doi. org/10. 1016/S 0140 -6736(17)31618 -5

Study Design

Objectives • The objective of ORAL Strategy was to directly compare the efficacy and safety of tofacitinib monotherapy, tofacitinib in combination with MTX and adalimumab in combination with MTX over 12 months in an adequately powered head-to-head trial, in patients with active rheumatoid arthritis and an inadequate response to previous MTX treatment. • Three independent, head-to-head treatment comparisons were conducted, with each arm directly compared with the other two: Ø Tofacitinib 5 mg BID + MTX vs adalimumab 40 mg Q 2 W + MTX Ø Tofacitinib 5 mg BID vs tofacitinib 5 mg BID + MTX BID=twice daily; MTX=methotrexate; Q 2 W=every 2 weeks. Fleischmann R et al. Lancet 2017; 390: 457 -68; Published online June 16, 2017; http: //dx. doi. org/10. 1016/S 0140 -6736(17)31618 -5

ORAL Strategy Study Design D -42 Randomization 1: 1: 1 MTX-IR Moderate. Severe RA Patients Screening A 1 -year, double-blind, triple-dummy, phase 3 b/4, active comparator, head-to-head, randomised controlled trial, assessing non-inferiority between treatment groups. Tofacitinib 5 mg BID + MTX Adalimumab 40 mg Q 2 W + MTX D 1 Patient Population ● Moderate to severe RA ● MTX-IR ● N=1152 (~380 per arm) HZ Vaccine l Eligible RA patients aged ≥ 50 years who were willing to volunteer received HZ vaccine 28 days prior to randomization Mo 6 (primary endpoint) Mo 12 Primary Endpoint ● ACR 50 at Mo 6 Comparisons (Noninferiority) ● Tofacitinib 5 mg BID + MTX vs ADA 40 mg Q 2 W + MTX ● Tofacitinib 5 mg BID vs tofacitinib 5 mg BID + MTX ACR 50=50% improvement in American College of Rheumatology response criteria; ADA=adalimumab; BID=twice daily; D=day; HZ=herpes zoster; IR=inadequate responder; Mo=month ; MTX=methotrexate; Q 2 W=every other week; RA=rheumatoid arthritis; N=Number. Fleischmann R et al. Lancet 2017; 390: 457 -68; Published online June 16, 2017; http: //dx. doi. org/10. 1016/S 0140 -6736(17)31618 -5

ORAL Strategy Patient Population Key Exclusion Criteria Key Inclusion Criteria Ø Moderate to severe RA – TJC/SJC ≥ 4 at screening and baseline/visit 2 (28 joint count) – hs. CRP ≥ 3 mg/L at screening – Score ≥ 6 on 2010 ACR/EULAR Classification Criteria Ø Moderate to severe RA inadequately controlled with MTX Ø Continuous MTX regimen >4 months prior to screening – Stable MTX dose 15– 25 mg/wk > 6 wks prior to BL Ø Ø Ø Contraindications for any study treatment; history of infections requiring treatment within 2 weeks, or any admission to hospital within the 6 months before randomization; had exclusionary morbidities, HIV, hepatitis B or C, inadequately treated or undocumented treatment of tuberculosis; more than one episode of herpes zoster, one episode of disseminated herpes zoster or herpes simplex; any clinically significant laboratory abnormalities, pregnant. Patients who had absence of efficacy or biological DMARD-related adverse events with previous treatment with a TNF inhibitor, or who had previously received tofacitinib, adalimumab, or glucocorticoids (equivalent to >10 mg per day prednisone within the previous 4 weeks), or live attenuated vaccines other than the herpes zoster vaccine (within 6 weeks before study initiation, or planned within 6 weeks after discontinuation of study treatment) were also excluded ACR=American College of Rheumatology; AE=adverse event; BL=baseline; EULAR=European League Against Rheumatism; hs. CRP=high sensitivity C-reactive protein; IR=inadequate responder; MTX=methotrexate; RA=rheumatoid arthritis; SJC=swollen joint count; TJC=tender joint count; TNF=tumor necrosis factor; wk=week. Fleischmann R et al. Lancet 2017; 390: 457 -68; Published online June 16, 2017; http: //dx. doi. org/10. 1016/S 0140 -6736(17)31618 -5

ORAL Strategy Endpoints Primary Efficacy Endpoint • ACR 50 response rates at Mo 6 testing for non-inferiority Safety Endpoints • All AEs, including serious AEs • Clinically significant abnormal laboratory parameter • Assessment of the rate of adverse reactions, injection site reactions, and development of zoster-like lesions after vaccination • Assessment of the rate of clinical HZ events by study treatment group after initiation of study treatments Key Secondary Efficacy Endpoints At Wk 6, Mo 3, 6, 9 and 12 • ACR 20 and ACR 70 response rates • Change from baseline in SDAI, CDAI • Change from baseline in DAS 28 -4(ESR) and DAS 28 -4(CRP) • Remission and low disease activity: SDAI, CDAI, DAS 28 -4(CRP), DAS 28 -4(ESR), and ACR-EULAR Boolean • Change from baseline in HAQ-DI • Change from baseline in SF-36 • Change from baseline in WPAI, EQ-5 D, FACIT-Fatigue scale ACR 20/50/70=20%, 50%, or 70% improvement in American College of Rheumatology response criteria; AE=adverse event; CDAI=clinical disease activity index; CRP=C-reactive protein; DAS 28=disease activity score of 28 joints; EQ-5 D=European quality of life 5 dimension questionnaire; ESR=erythrocyte sedimentation rate; EULAR=European League Against Rheumatism; FACIT=Functional Assessment of Chronic Illness Therapy; HAQ-DI=Health Assessment Questionnaire-Disability Index; HZ=herpes zoster; Mo=month; SDAI=simplified disease activity index; SF-36=36 item Short-Form Health Survey; wk=week; WPAI=work productivity and activity impairment

Statistical Methodology

Non-Inferiority Clinical Trials to Establish Effectiveness: FDA Guidance for Industry https: //www. fda. gov/downloads/drugs/guidancecomplianceregulatoryinformation/guidances/ucm 202140. pdf

Reporting of Non-inferiority Trials 0 A B C Superior The lower bound of CI is greater than 0 Non-inferior The lower bound of CI is greater than Non-inferior Not demonstrated The lower bound of CI is less than D 1 Inferior D 2 The upper bound of CI is less than 0 (NI margin) Favor Control Drug 0 Treatment difference CI=confidence interval; NI=non-inferiority. Figure adapted from: Piaggio G et al. JAMA 2006; 295: 1152 -1160. Favor Test Drug

ORAL Strategy Study Design: Non-inferiority Margin • Incorporated practical guidance from the FDA for non-inferiority studies 1 • Per standard practice, a non-inferiority margin of ~50% of the observed treatment difference in recent trials of adalimumab + MTX vs placebo was calculated 2, 3 • The margin was found to be − 13%4 • Bonferroni corrections suggested an alpha value of 0. 0166; therefore non-inferiority was confirmed if the lower bound of the 98. 34% CI was greater than − 13% • If non-inferiority was shown, superiority was declared if the lower bound of the 98. 34% CI of the difference was greater than zero 1. https: //www. fda. gov/downloads/drugs/guidancecomplianceregulatoryinformation/guidances/ucm 202140. pdf 2. Machado MA et al. Rev Bras Reumatol 2013; 53: 419 -430 3. van Vollenhoven RF et al. N Engl J Med 2012; 367: 508 -519 4. Fleischmann R et al. Lancet 2017; 390: 457 -68; Published online June 16, 2017; http: //dx. doi. org/10. 1016/S 0140 -6736(17)31618 -5

Statistical Methods: Primary Analysis • Primary Endpoint: ACR 50 at Mo 6 • Three independent treatment comparisons – Tofacitinib + MTX vs adalimumab + MTX – Tofacitinib monotherapy vs tofacitinib +MTX • 98. 34% 2 -sided CIs will be constructed for the treatment differences (with Bonferroni correction) • Based on Full Analysis Set (modified ITT), with sensitivity analysis based on Per-Protocol Set ACR 50=50% improvement in American College of Rheumatology response criteria; ADA=adalimumab; CI=confidence interval; ITT=intent-to-treat; mo=month; MTX=methotrexate. 28

Patient Demographics

ORAL Strategy: Subject Population 1803 patients screened 1152 patients randomized 386 patients assigned to tofacitinib 5 mg monotherapy 378 patients assigned to tofacitinib 5 mg + MTX 388 patients assigned to adalimumab 40 mg + MTX 69 patients discontinued 73 patients discontinued 74 patients discontinued 315 patients completed the study 303 patients completed the study 312 patients completed the study FAS: 384 patients FAS: 376 patients FAS: 386 patients PPAS: 338 (88. 0%) patients PPAS: 320 (85. 1%) patients PPAS: 330 (85. 5%) patients 6 subjects were randomized but did not receive any study drug. These patients are not included in the FAS or safety analysis FAS=full analysis set; MTX=methotrexate; PPAS=per-protocol analysis set.

ORAL Strategy: Reasons for Discontinuation Tofacitinib 5 mg BID N=386 n (%) Tofacitinib 5 mg BID + MTX N=378 n (%) Adalimumab 40 mg + MTX N=388 n (%) 69 (17. 9) 73 (19. 3) 74 (19. 1) Due to death 2 (0. 5) 0 0 Due to AE 21 (5. 5) 26 (6. 9) 36 (9. 3) Due to insufficient clinical response 10 (2. 6) 3 (0. 8) 7 (1. 8) Lost to follow-up 2 (0. 5) 7 (1. 9) 3 (0. 8) 0 2 (0. 5) 0 No longer willing to participate in study 6 (1. 6) 8 (2. 1) 10 (2. 6) Protocol violation 9 (2. 3) 9 (2. 4) 4 (1. 0) Withdrawn due to pregnancy 3 (0. 8) 1 (0. 3) 0 Withdrew consent 11 (2. 9) 11 (2. 8) Other 5 (1. 3) 6 (1. 6) 3 (0. 8) Discontinued No longer meets eligibility criteria AE=adverse event; BID=twice daily; MTX=methotrexate.

Baseline Demographics (FAS) Tofacitinib 5 mg BID monotherapy (N=384) Tofacitinib 5 mg BID + MTX (N=376) Adalimumab 40 mg Q 2 W + MTX (N=386) Gender, female (%) 319 (83) 311 (83) 320 (83) Gender, male (%) 65 (17) 66 (17) 49. 7 (12. 2) 50. 0 (13. 4) 50. 7 (13. 4) Race, white (%) 296 (77) 286 (76) 293 (76) Race, black (%) 11 (3) 19 (5) 18 (5) Race, asian (%) 41 (11) 38 (10) 40 (10) Race, other (%) 36 (9) 33 (9) 35 (9) 6. 1 (0. 2– 41. 6) 5. 4 (0. 0– 43. 5) 6. 0 (0. 3– 42. 8) Prior cs. DMARD (excl. MTX), n (%) 122 (32) 115 (31) 142 (37) Prior b. DMARD (excl. TNFi), n (%) 17 (4) 14 (4) 20 (5) Prior TNFi, n (%)* 28 (7) 16 (4) 19 (5) Corticosteroid use at baseline, n (%) 223 (58) 214 (57) 218 (57) Mean corticosteroid dose at baseline, mg/d (SD) 7. 3 (13. 3) 6. 5 (2. 5) 6. 5 (2. 6) 0 16. 7 (3. 7) 16. 4 (3. 7) Age (years), mean (SD) Disease duration (years), median (range) Background weekly MTX dose, mg (SD) *Patients who had previously had an inadequate response to TNFi therapy were excluded b. DMARD=biologic disease-modifying antirheumatic drug; BID=twice daily; cs. DMARD=conventional synthetic disease-modifying antirheumatic drug; FAS=full analysis set; MTX=methotrexate; Q 2 W=every 2 weeks; SD=standard deviation; TFNi=tumor necrosis factor inhibitor. Fleischmann R et al. Lancet 2017; 390: 457 -68; Published online June 16, 2017; http: //dx. doi. org/10. 1016/S 0140 -6736(17)31618 -5.

Baseline Disease Characteristics (FAS) Tofacitinib 5 mg BID monotherapy (N=384) Tofacitinib 5 mg BID + MTX (N=376) Adalimumab 40 mg Q 2 W + MTX (N=386) TJC (28), mean (SD) 15. 4 (6. 5) 15. 6 (6. 5) 15. 2 (6. 7) SJC (28), mean (SD) 11. 2 (5. 6) 11. 8 (5. 7) 11. 0 (5. 4) Pt. GA, mean (SD) 60. 1 (21. 4) 61. 7 (22. 0) 60. 2 (23. 5) PGA, mean (SD) 59. 7 (17. 7) 60. 7 (18. 0) 60. 3 (19. 6) Pain (VAS 1– 100), mean (SD) 61. 2 (21. 7) 60. 7 (22. 4) 60. 6 (22. 6) SDAI, mean (SD) 40. 2 (13. 0) 41. 6 (13. 2) 39. 8 (13. 3) CDAI, mean (SD) 38. 6 (12. 6) 39. 7 (12. 7) 38. 2 (12. 9) DAS 28 -4(ESR), mean (SD) 6. 5 (0. 9) 6. 6 (0. 9) 6. 5 (1. 0) DAS 28 -4(CRP), mean (SD) 5. 7 (0. 9) 5. 8 (0. 9) 5. 7 (1. 0) HAQ-DI, mean (SD) 1. 6 (0. 6) 16. 6 (19. 3) 18. 7 (21. 9) 16. 7 (21. 3) hs. CRP (mg/L), mean (SD) Pain (VAS), patient assessment of arthritis pain (VAS) BID=twice daily; CDAI=clinical disease activity index; DAS=disease activity score; ESR=erythrocyte sedimentation rate; FAS, full analysis set; HAQ-DI=Health Assessment Questionnaire-Disability Index; hs. CRP=high sensitivity C-reactive protein; MTX=methotrexate; PGA=Physician Global Assessment; Pt. GA= Patient Global Assessment; SD=standard deviation; TFNi=tumor necrosis factor inhibitor; SJC=swollen joint count; SD=standard deviation; SDAI=simple disease activity index; TJC=tender joint count; VAS=visual analogue scale. Fleischmann R et al. Lancet 2017; 390: 457 -68; Published online June 16, 2017; http: //dx. doi. org/10. 1016/S 01406736(17)31618 -5

Efficacy Outcomes

Analysis of Primary Endpoint (Non-inferiority of ACR 50 at 6 Months) (FAS) Difference and 98. 34% CI between groups in ACR 50 response rates at Month 6 (FAS) Non-inferiority margin (− 13%) 2. 2% * Tofacitinib 5 mg BID + MTX vs Adalimumab + MTX -6. 4% 10. 9% -5. 5% 3. 0% -14. 0% Tofacitinib 5 mg BID vs Adalimumab + MTX -7. 7% Tofacitinib 5 mg BID vs Tofacitinib 5 mg BID + MTX -16. 3% -20 -15 0. 8% -10 -5 0 5 10 Non-inferior *Criteria for non-inferiority met Non-inferiority not demonstrated Inferior Superior » Similar results were seen with the per protocol sensitivity analysis Fleischmann R et al. Lancet 2017; 390: 457 -68; Published online June 16, 2017; http: //dx. doi. org/10. 1016/S 0140 -6736(17)31618 -5 15

Tofacitinib + MTX was Non-inferior to Adalimumab + MTX as Measured by ACR 50 at Month 6 Tofacitinib 5 mg BID (n=384) Tofacitinib 5 mg BID + MTX (n=376) Response Rate (%) FAS 100 ADA 40 mg + MTX (n=386) -5. 50 (-14. 0, 3. 0) 80 73 65 71 -7. 7 (-16. 3, 0. 8 2. 2 (- 6. 4, 10. 9) 60 46 38 40 44 25 18 20 0 ACR 20 • • • ACR 50 ACR 70 Similar ACR 50 response rate at month 6 in MTX-IR Tofacitinib as a single agent did not meet non-inferiority criteria compared with Tofacitinib + MTX Tofacitinib as a single agent did not meet non-inferiority criteria compared with Adalimumab + MTX Primary endpoint =ACR 50 response rate at 6 months Secondary endpoints = ACR 20, ACR 70 at 6 months ACR 20/50/70=20%, 50%, or 70% improvement in American College of Rheumatology response criteria; ADA=adalimumab; BID=twice daily; FAS=full analysis set; IR=inadequate responder; MTX=methotrexate; Fleischmann R et al. Lancet 2017; 390: 457 -68; Published online June 16, 2017; http: //dx. doi. org/10. 1016/S 0140 -6736(17)31618 -5 21

ACR 20, ACR 50, and ACR 70 Response Rates ACR 20 Response Rate (%) 100 ACR 50 100 80 80 80 60 60 60 40 40 40 20 20 20 0 BL Wk 6 Mo 3 0 Mo 6 Mo 9 Mo 12 ACR 70 100 0 BL Wk 6 Mo 3 Mo 6 Mo 9 Mo 12 BL Wk 6 Mo 3 ACR 20/50/70=20%, 50%, or 70% improvement in American College of Rheumatology response criteria; BID=twice daily; BL=baseline; mo=month; MTX=methotrexate; wk=week. Mo 6 Mo 9 Mo 12 37

SDAI and CDAI Change From Baseline 0 0 SDAI -5 Mean Change From Baseline CDAI -10 -15 -20 -25 -30 -35 BL Wk 6 Mo 3 Mo 6 Mo 9 Mo 12 Baseline SDAI: Tofacitinib 5 mg BID=40. 2; Tofacitinib 5 mg BID + MTX=41. 6; adalimumab 40 mg + MTX=39. 8. Baseline CDAI: Tofacitinib 5 mg BID=38. 6; Tofacitinib 5 mg BID + MTX=39. 7; adalimumab 40 mg + MTX=38. 2. BID=twice daily; BL=baseline; CDAI=clinical disease activity index; mo=month; MTX=methotrexate; SDAI=simplified disease activity index; wk=week. 38

DAS 28 -4(ESR) Change From Baseline DAS 28 -4(ESR) 0. 0 Mean Change From Baseline -0. 5 -1. 0 -1. 5 -2. 0 -2. 5 -3. 0 BL Mo 3 Mo 6 Mo 12 Baseline DAS 28(ESR): Tofacitinib 5 mg BID=6. 5; Tofacitinib 5 mg BID + MTX=6. 6; adalimumab 40 mg + MTX=6. 5. BID=twice daily; BL=baseline; DAS 28=disease activity score of 28 joints; ESR=erythrocyte sedimentation rate; mo=month; MTX=methotrexate; wk=week. 39

Rate of Remission Rate (%) 80 60 40 20 0 Remission Rate (%) 100 SDAI≤ 3. 3 BL 100 Wk 6 Mo 3 Mo 6 DAS 28 -4(ESR)<2. 6 80 60 * 40 20 0 BL Mo 3 Mo 6 Mo 12 CDAI≤ 2. 8 80 60 40 20 0 Mo 12 Remission Rate (%) 100 BL Wk 6 Mo 3 Mo 6 Mo 9 Mo 12 ACR-EULAR Boolean 100 80 60 40 20 0 BL Wk 6 Mo 3 Mo 6 Mo 9 ACR=American College of Rheumatology; BID=twice daily; BL=baseline; CDAI=clinical disease activity index; DAS 28=disease activity score of 28 joints; ESR=erythrocyte sedimentation rate; EULAR=European League Against Rheumatism; mo=month; MTX=methotrexate; SDAI=simple disease activity index; wk=week. Mo 12 40

100 SDAI ≤ 11 80 60 60 40 40 20 0 CDAI ≤ 10 80 20 BL Wk 6 Mo 3 Mo 6 Response Rate (%) 100 Response Rate (%) Low Disease Activity (LDA) Mo 9 100 0 Mo 12 BL Mo 3 Mo 6 Mo 12 DAS 28 -4(ESR) ≤ 3. 2 80 60 40 20 0 BL Mo 3 Mo 6 Mo 12 ACR=American College of Rheumatology; BID=twice daily; BL=baseline; CDAI=clinical disease activity index; DAS 28=disease activity score of 28 joints; ESR=erythrocyte sedimentation rate; EULAR=European League Against Rheumatism; mo=month; MTX=methotrexate; SDAI=simplified disease activity index; wk=week. 41

Analysis of Change From Baseline in HAQ-DI Mean Change From Baseline 0. 0 -0. 1 MCID=-0. 22 -0. 3 -0. 4 -0. 5 -0. 6 -0. 7 BL Wk 6 Mo 3 Mo 6 Mo 9 Mo 12 Baseline HAQ-DI: Tofacitinib 5 mg BID=1. 6; Tofacitinib 5 mg BID + MTX=1. 6; adalimumab 40 mg + MTX=1. 6. BID=twice daily; BL=baseline; HAQ-DI=Health Assessment Questionnaire Disability Index; MCID=minimal clinically important difference; mo=month; MTX=methotrexate; wk=week. 42

Analysis of Change From Baseline in the SF-36 Physical Component Score Mean Change from Baseline (SD) 10 9 8 7 6 5 4 3 MCID=2. 5 2 1 0 BL Mo 3 Mo 6 Mo 9 BID=twice daily; BL=baseline; MCID=minimal clinically important difference; mo=month; MTX=methotrexate; SD=standard deviation; SF-36=36 item Short-Form Health Survey. Mo 12 43

Analysis of Change From Baseline in the SF-36 Mental Component Score Mean Change From Baseline (SD) 7 6 5 4 MCID=2. 5 3 2 1 0 BL Mo 3 Mo 6 Mo 9 BID=twice daily; BL=baseline; MCID=minimal clinically important difference; mo=month; MTX=methotrexate; SD=standard deviation; SF-36=36 item Short-Form Health Survey. Mo 12 44

Analysis of Change From Baseline in the FACIT-Fatigue Scale Mean Change From Baseline (SD) 9 8 7 6 5 MCID=4 4 3 2 1 0 BL Mo 3 Mo 6 Mo 9 BID=twice daily; BL=baseline; FACIT=functional assessment of chronic illness therapy; MCID=minimal clinically important difference; mo=month; MTX=methotrexate; SD=standard deviation. Mo 12 45

Safety Outcomes

Safety: AEs, SAEs and Deaths (SAS) Tofacitinib 5 mg BID monotherapy (N=384) Tofacitinib 5 mg BID + MTX (N=376) Adalimumab 40 mg Q 2 W + MTX (N=386) Total number of AEs, n* 598 652 620 Patients with AEs, n (%) 226 (59) 231 (61) 253 (66) Patients with treatment-related AEs, n (%) 101 (26) 111 (30) 133 (35) Patients with serious AEs, n (%) 35 (9) 27 (7) 24 (6) Patients discontinuing due to AEs, n (%) 23 (6) 26 (7) 37 (10) Patients with severe AEs, n (%) 24 (6) 17 (5) 23 (6) Deaths, n (%)† 2 (1) 0 0 *Patients could have experienced more than one AE (n=1); atypical pneumonia and respiratory distress syndrome associated with influenza A (n=1) †Urosepsis AE=adverse event; BID=twice daily; BL=baseline; MTX=methotrexate; Q 2 W=every 2 weeks; SAE=serious adverse event; SAS=safety analysis set. Fleischmann R et al. Lancet 2017; 390: 457 -68; Published online June 16, 2017; http: //dx. doi. org/10. 1016/S 0140 -6736(17)31618 -5

Safety: AEs of Special Interest (SAS) Tofacitinib 5 mg BID monotherapy (N=384) Tofacitinib 5 mg BID + MTX (N=376) Adalimumab 40 mg Q 2 W + MTX (N=386) Serious infections, n (%) 6 (2) 10 (3) 6 (2) Herpes zoster (serious and non-serious), n (%) 4 (1) 8 (2) 6 (2) Opportunistic infections (excl. tuberculosis), n (%) a 2 (1) 1 (<1) 2 (1) 0 0 0 2 (1) 1 (<1%) 0 0 2 (1) 0 1 (<1) 110 (29) 17 (5) 164 (44) 53 (14) 182 (47) 62 (16) 7 (2) 29 (8) 27 (7) 85 (22 11 (3) 3 (<1) 129 (34) 33 (9) 16 (4) 143 (37) 38 (10) 15 (4) 6 (2) 1 (<1) 15 (4) 0 0 Tuberculosis, n (%) MACE (non-fatal), n (%) Malignancy (excl. NMSC), n (%) b NMSC, n (%) Liver function test abnormalities, n (%) ALT ≥ 1 x ULN ALT ≥ 2 x ULN ALT ≥ 3 x ULN AST ≥ 1 x ULN AST ≥ 2 x ULN AST ≥ 3 x ULN Total bilirubin ≥ 1 ≥ 2 ≥ 3 a. Tofacitinib 5 mg BID: herpes zoster (n=2); tofacitinib 5 mg BID + MTX: cytomegalovirus chorioretinitis (n=1): adalimumab 40 mg Q 2 W + MTX: herpes zoster (n=1), herpes varicella (n=1) b. Tofacitinib 5 mg BID: breast cancer (n=1) ALT=alanine aminotransferase; AST=aspartate aminotransferase; BID=twice daily; MACE=major adverse cardiovascular events; MTX=methotrexate; NMSC=non-melanoma skin cancer; Q 2 W=every 2 weeks; SAS=safety analysis set; ULN=upper limit of normal. Fleischmann R et al. Lancet 2017; 390: 457 -68; Published online June 16, 2017; http: //dx. doi. org/10. 1016/S 0140 -6736(17)31618 -5

Conclusions • For ACR 50 response rates at 6 months, the combination of tofacitinib + MTX was found to be non-inferior to adalimumab + MTX 1 – Although tofacitinib monotherapy was effective (and similar to observations for b. DMARDs + MTX), 2 non-inferiority was not shown for tofacitinib monotherapy vs either combination arm 1 • These results suggest that in patients with an inadequate response to methotrexate, the addition of tofacitinib or adalimumab is equally efficacious and more effective than switching to tofacitinib monotherapy 1. Fleischmann R et al. Lancet 2017; 390: 457 -68; Published online June 16, 2017; http: //dx. doi. org/10. 1016/S 0140 -6736(17)31618 -5 2. Smolen JS et al. Lancet 2016; 388: 2023 -2038.