Briefing Document for the Joint Meeting of the Arthritis Advisory Committee and the Drug Safety and Risk Management Advisory Committee 11 January 2019 Uloric (febuxostat) NDA 021856 Sponsor: Takeda Pharmaceuticals USA One Takeda Parkway Deerfield, Illinois 60015 Advisory Committee Briefing Materials: Available for Public Release Uloric (Febuxostat) Page 2of 82 AdvisoryCommittee Briefing Document 11 January 2019 1.0 EXECUTIVE SUMMARY Febuxostat is a potent, nonpurine, selective xanthine oxidase inhibitor (XOI). It was approved on 13 February 2009 in the United States (US) at doses of 40 and 80 mg once daily (QD) for the chronic management of hyperuricemia in patients with gout. It is not recommended for treatment of asymptomatic hyperuricemia in the US. It was initially approved in the European Union (EU) (international birth date [IBD] 21 April 2008), and febuxostat has since been approved in 85 countries worldwide. In some ex-US countries, febuxostat is also indicated for the treatment of hyperuricemia in patients without gout and/or prevention and treatment of hyperuricemia in adult patients undergoing chemotherapy for hematologic malignancies at intermediate-to-high risk of tumor lysis syndrome. Approved doses vary in different regions and range from 10to 120 mg. The estimated global exposure is 15.1 million person-years (PY); the total estimated US exposure is 1.4 million PY. Gout has an estimated prevalence of 8 to 9 million adults in the US [1]. Gout is characterized by flares of acute arthritis, chronic gouty arthropathy, tophi, and uric acid urolithiasis, and is associated with a broad range of comorbidities, including cardiovascular (CV) disease, chronic kidney disease (CKD), and metabolic syndrome [2-4]. The most frequent initial clinical manifestation of gout is a flare of inflammatory arthritis, which develops over a few hours, usually involves a single lower extremity(LE) joint, and is severely disabling due to extreme pain, swelling, and loss of joint function. In most patients, the initial attack resolves spontaneously and completely within a week or 2, but in most instances, recurs intermittently and progressively more often. The underlying metabolic aberration in gout is hyperuricemia. It is defined as an elevation in plasma or serum uric acid (sUA) ≥6.8 mg/dL, the limit of urate solubility, which signals a risk for gout through crystal formation and deposition. Over time, the crystal deposition and inflammation can result in progressive joint damage and erosion, with the formation of masses of urate crystals in a chronic inflammatory background (called tophi) that can be deforming and disabling. Patients who experience frequent gout flares or have tophi present have a lower quality of life [5]. Adequate control of gout not only reduces the physical manifestations of the disease, it also reduces associated healthcare burdens. Treatment and prevention of gout traditionally involved anti-inflammatory medication for acute attacks (eg, colchicine, corticosteroids or nonsteroidal anti-inflammatory drugs [NSAIDs]), and urate-lowering therapy (ULT) for long-term management of hyperuricemia and prevention of urate crystal-induced damage. To achieve the optimum treatment goals for gout, the American College of Rheumatology (ACR) gout guidelines recommends that a ULT include a target serum urate goal of <6.0 mg/dL for most gout patients, with a goal of <5.0 mg/dL in patients with tophaceous disease or more severe gout [6]. Overall, there are few ULTs available and each agent has limitations. ULTs can be grouped into 3 categories based on their mechanism of action:  XOIs: reduce uric acid production (febuxostat, allopurinol).  Uricosurics: promote urinary uric acid clearance (probenecid, lesinurad). Advisory Committee Briefing Materials: Available for Public Release Uloric (Febuxostat) Page 3of 82 AdvisoryCommittee Briefing Document 11 January 2019  Recombinant uricase: a biological agent that enzymatically degrades circulating uric acid (pegloticase). XOIs are the consensus first-line ULT, a role most often fulfilled with allopurinol, or alternatively febuxostat [6]. The clinical development program for febuxostat supporting approval is shown in Figure 1.a. Figure 1.a Febuxostat Clinical Development Program Supporting Approval Study 004 (phase 2), APEX, FACT, and CONFIRMS (phase 3) were randomized, double-blind studies. FOCUS and EXCEL were open-label extension studies. Bothfebuxostat 40 mg and 80 mg doses demonstrated efficacy in urate lowering in subjects with gout. Febuxostat demonstrated the ability to achieve sUA targets of <6.0 and/or <5.0 mg/dL better than commonly prescribed doses (≤300 mg/day) of allopurinol in current clinical practice, including in subjects with sUA levels ≥10 mg/dL or who have tophi. Additionally, febuxostat decreases acute flares and reduces tophus size over time through persistent long-term urate- lowering effects. More importantly, unlike allopurinol, febuxostat has been approved without dose adjustment for patients with gout and mild-to-moderate renal impairment (estimated creatinine clearance [eCrCl] of 30 to 90 mL/min), and febuxostat 40 mg is also approved in patients with severe renal impairment (eCrClof <30 mL/min). In the initial 2 phase 3 studies (APEX and FACT), Antiplatelet Trialists’ Collaborative (APTC) events (CV death, nonfatal myocardial infarction [MI], nonfatal stroke) were numerically low, but there was an imbalance in the rate of events for febuxostat 80 and 120 mg compared with allopurinol or placebo (Table 5.b). As a result, Takeda conducted an additional phase 3 study (CONFIRMS) that assessed safety and efficacy of the febuxostat 40 and 80 mg doses compared with allopurinol in subjects with hyperuricemia and gout. CONFIRMS did not show a higher rate of CV thromboembolic events with febuxostat than with allopurinol (Table 5.c). A pooled analysis of the randomized phase 3 studies revealed a numerical imbalance in the number of CV thromboembolic events in the composite of CV deaths, nonfatal MI, and nonfatal strokes in subjects treated with febuxostat compared with allopurinol as summarized in Table 1.a. Advisory Committee Briefing Materials: Available for Public Release Uloric (Febuxostat) Page 4of 82 AdvisoryCommittee Briefing Document 11 January 2019 Table 1.a Adjudicated Major CV Events From Pooled Analysis of Phase 3 Randomized Controlled Studies (APEX, FACT, CONFIRMS) Events, n (rate per 100 subject years) 95% CI Febuxostat Allopurinol N=2690 N=1277 All APTC events 10 (0.74) 4 (0.60) 0.36, 1.37 0.16,1.53 CV death 3 (0.22) 2 (0.30) 0.05, 0.65 0.04, 1.08 Nonfatal MI 5 (0.37) 2 (0.30) 0.12, 0.87 0.04, 1.08 Nonfatal stroke 2 (0.15) 0 0.02, 0.54 0.00, 0.55 APTC: Antiplatelet Trialists’ Collaborative; CV: cardiovascular; MI: myocardial infarction. Adjudication was done prospectively in CONFIRMS and retrospectively in APEX and FACT. These data were reviewed by the Food and Drug Administration (FDA) Arthritis Advisory Committee in November 2008, where based on the evidence presented the committee voted 12 to 0 recommending approval with 1 abstention. The committee highlighted that there was adequate evidence of the efficacy and safety of febuxostat. However, the advisory committee agreed that the CV safety risk had not been fully elucidated and a postapproval study should be conducted to further assess the CV safety of the drug. Febuxostat was approved with a Warnings and Precautions for Cardiovascular Events in the prescribing information highlighting the rate of CV events from the pooled analysis and a postmarketing requirement to conduct a randomized controlled study of adequate size and duration to determine whether febuxostat was associated with an increase in the risk of serious adverse CV outcomes as compared with allopurinol[7]. The CARES study was conducted to fulfill this requirement, and was designed to assess if febuxostat was noninferior to allopurinol for the primary endpoint of the major adverse cardiac event (MACE) composite (CV death, nonfatal MI, nonfatal stroke, and unstable angina with urgent coronary revascularization) in subjects with gout and a high CV risk profile [8]. The study had a prespecified noninferiority margin of 1.3 for the hazard ratio (HR) for the primary endpoint. The study protocol was finalized following discussions and agreement with FDA. The study was initiated in 2010 and completed in 2017. Subjects with major concurrent CV disease were randomly assigned to receive febuxostat or allopurinol and stratified by kidney function. Doses were titrated to achieve target sUA <6 mg/dL. Unlike febuxostat, doses of allopurinol were titrated based on kidney function. A total of 6190 subjects were randomized into the study and were followed for a median of 32 months. The cumulative duration of study participation was similar between treatment groups. Baseline and demographic characteristics were similar between treatment groups, and most subjects were Advisory Committee Briefing Materials: Available for Public Release Uloric (Febuxostat) Page 5of 82 AdvisoryCommittee Briefing Document 11 January 2019 male and white; the overall mean age was 64.8 years. Over half in each treatment group had moderate renal impairment (eCrCl 30 to 59 mL/min). The CARES study results are shown in Table 1.b.  For the primary endpoint of MACE composite, CARES showed that febuxostat was noninferior to allopurinol.  When analyzing the individual components of MACE as secondary endpoints, the rate ofCV death was higher with febuxostat compared with allopurinol whereas the individual rates for nonfatal MI, nonfatal stroke, and urgent revascularization for unstable angina did not differ between febuxostat and allopurinol. All-cause mortality was also higher with febuxostat than allopurinol due to the higher rate of CV deaths [9]. Table 1.b Analysis of MACE and Each Event in the MACE Composite (mITT) Febuxostat Allopurinol N =3098 N =3092 Hazard ratio (95% CI)a Composite primary endpoint 335 (10.8) 321 (10.4) 1.03 (0.87, 1.23)b CV death 134 (4.3) 100 (3.2) 1.34 (1.03,1.73) Nonfatal MI 111 (3.6) 118 (3.8) 0.93 (0.72, 1.21) Nonfatal stroke 71 (2.3) 70 (2.3) 1.01 (0.73, 1.41) Unstable angina with urgent coronary revascularization 49 (1.6) 56 (1.8) 0.86 (0.59, 1.26) Source: White et al [9]. CV: cardiovascular; MACE: major adverse cardiac event; MI: myocardial infarction; mITT: modified intent-to-treat. Analysis based on mITT. aFebuxostat to allopurinol. b95% adjusted CI.  Because ofthe higher rate of CV death, multiple potential risk factors were evaluated including: subject baseline characteristics, changes in electrolytes, changes in blood pressure or heart rate, and gout flare rates. Overall, there was no biologically plausible explanation for the higher rate of CV death.  The design of the CARES study has both strengths and limitations: – CARES was the first cardiovascular outcomes trial (CVOT) of urate lowering drugs in subjects with gout and major CV or cerebrovascular disease. It wasthe largest, long-term, double-blind, randomized study that incorporated a predefined, comprehensive adjudication process for major CV endpoints to address the question of CV risk of febuxostat compared with allopurinol. Moreover, the study allowed for direct comparison of allopurinol and febuxostat using a treat-to-target dose-titration scheme. – The CARES study is limited by the lack of a placebo arm, which precludes determination of either study drugs’ impact in CV risk compared withan untreated gout population. Additionally, because many subjects enrolled had previous exposure to allopurinol, this Advisory Committee Briefing Materials: Available for Public Release Uloric (Febuxostat) Page 6of 82 AdvisoryCommittee Briefing Document 11 January 2019 may create a potential bias for underestimation of adverse events with allopurinol, as subjects who previously did not tolerate allopurinol would not have enrolled. The CARES study was not designed for noninferiority for the individual MACE component endpoints. Furthermore, type 1 error control and noninferiority assessment were done solely for the primary endpoint and not the individual MACE components. – In this study, there were many subjects who discontinued from study drug and many subjects who did not complete follow-up. High rates of discontinuation of treatment can lead to concerns of missing a significant difference between treatment groups in the primary or secondary outcomes. Analyses were conducted to assess the potential impact of missing data including overall dropout rates, reason for dropout, timing of dropout, baseline characteristics of subjects who dropped out, and multiple imputation. These additional analyses were consistent with the modified intent-to-treat (mITT) analyses.  There was no signal for increased CV risk for febuxostat from nonclinical data or a thorough correctedQT interval(QTc) study conducted in healthy subjects. Postmarketing signal detection also revealed no signal for CV events or CV death with febuxostat. Published studies in the literature have varied results and do not support an increased CV risk with febuxostat. Further understanding of the CV profile for febuxostat and allopurinol may be gained from an on-going CV safety study being conducted in Europe as a postauthorization requirement by Menarini (a partner of Teijin Pharmaceutical Companies, Takeda’s alliance partner) the Febuxostat versus Allopurinol Streamlined Trial (FAST) [10]. It is a prospective, randomized, parallel group, open-label, blinded endpoint (PROBE) study comparing febuxostat with allopurinol conducted in the clinical setting (general practitioners and specialist). It is designed as a noninferiority study and the primary endpoint is first occurrence of the APTC CV endpoint of nonfatal MI, nonfatal stroke,or CV death. The recruitment in FAST ended in January 2018 and the clinical study report is expected by August 2020. Following comprehensive clinical assessments and exploratory analyses, Takeda has been unable to determine any contributing factors or a population at risk for the higher rate of CV death observed in the CARES study. Accounting for the totality of the evidence, including the lack of concordance between the individual MACE components, the inability to explain the finding and the absence of biological plausibility, there remains uncertaintyabout the reliability of the observation. The seriousness of the observation in CARES makes it important that physicians are informed so they can consider the benefits and risks of febuxostat compared with other treatment options for each individual patient based upon the best available information. The CARES data has been widely disseminated to the medical community over the past several months including publication of CARES in the New England Journal of Medicine in March 2018 [9]. Takeda will continue to share the CARES study findings with the medical community. In addition, Takeda proposes to update the prescribing information with the CARES study data to include CV death in the Warnings and Precautions section and describe the CARES study Advisory Committee Briefing Materials: Available for Public Release Uloric (Febuxostat) Page 7of 82 AdvisoryCommittee Briefing Document 11 January 2019 results in the clinical trials section and all other relevant sections. Specific labeling language will be discussed with FDA during labeling review. Takeda will communicate these important updates following label approval, via distribution of a Dear Healthcare Provider (HCP) letter to HCPs, pharmacies, and professional societies. Febuxostat continues to be an important treatment option in the management of hyperuricemia in patients with gout. Considering the seriousness of the disease, the limited treatment options, the clinical use of febuxostat over 10 years and its safety profile, including the totality of the CV safety data, the benefit-risk assessment for febuxostat remains positive. Advisory Committee Briefing Materials: Available for Public Release Uloric (Febuxostat) Page 8of 82 AdvisoryCommittee Briefing Document 11 January 2019 TABLE OF CONTENTS 1.0 EXECUTIVE SUMMARY..............................................................................................2 2.0 BACKGROUND AND REGULATORY HISTORY.....................................................14 3.0 GOUT OVERVIEW......................................................................................................17 3.1 Gout Disease Background........................................................................................17 3.2 Cardiorenal Comorbidities.......................................................................................17 3.3 Treatment Landscape...............................................................................................19 4.0 CARES STUDY............................................................................................................24 4.1 Overview.................................................................................................................24 4.2 Study Design...........................................................................................................24 4.3 Governance Structure...............................................................................................26 4.3.1 Data Monitoring Committee...............................................................................26 4.3.2 CEC...................................................................................................................26 4.4 Statistical Analysis...................................................................................................27 4.5 Interim Analyses......................................................................................................28 4.6 Overall Exposure and Disposition............................................................................29 4.6.1 Exposure............................................................................................................29 4.6.2 Final Titrated Dose............................................................................................29 4.6.3 Disposition.........................................................................................................30 4.6.4 Subject Retention Measures...............................................................................31 4.6.5 Collection of Vital Status...................................................................................31 4.7 Demographic and Other Baseline Characteristics and Concomitant Medications......32 4.8 CV Outcomes..........................................................................................................32 4.8.1 Primary Endpoint: MACE Composite................................................................32 4.8.2 Secondary Endpoint: APTC Composite (CV death, Nonfatal MI, and Nonfatal Stroke)...............................................................................................................37 4.8.3 Secondary Endpoint: Analysis of Time to Occurrence of Each Event in the MACE Composite..............................................................................................37 4.8.4 Analysis of All-Cause Mortality.........................................................................43 4.8.5 Exploratory Analyses to Understand the CV Death Data....................................44 4.8.6 Assessing the Impact of Subject Discontinuation................................................49 4.9 Summary of Non-CV Safety....................................................................................51 4.10 Strengths and Limitations of the CARES Study.......................................................52 5.0 PRECLINICAL AND OTHER CLINICAL DATA RELEVANT TO CV SAFETY......54 5.1 Preclinical Program..................................................................................................54 Advisory Committee Briefing Materials: Available for Public Release Uloric (Febuxostat) Page 9of 82 AdvisoryCommittee Briefing Document 11 January 2019 5.2 Clinical Development Program................................................................................54 5.2.1 Phase 1 Study.....................................................................................................54 5.2.2 Phase 3 Development Program...........................................................................55 5.3 Other CV Safety Study: FAST.................................................................................56 6.0 POSTMARKETING SURVEILLANCE.......................................................................58 7.0 LITERATURE REVIEW..............................................................................................59 8.0 ASSESSMENT OF CV SAFETY OF FEBUXOSTAT..................................................62 8.1 CARES....................................................................................................................62 8.2 Development Program.............................................................................................63 8.3 Other Relevant Evidence..........................................................................................63 8.4 Overall CV Safety Summary....................................................................................63 9.0 FEBUXOSTAT EFFICACY.........................................................................................64 9.1 Clinical Efficacy in the Febuxostat Development Program.......................................64 9.2 Clinical Efficacy in CARES.....................................................................................67 9.2.1 sUA...................................................................................................................67 9.2.2 Gout Flare..........................................................................................................68 9.2.3 Tophi.................................................................................................................69 9.3 Additional Efficacy Data From a Febuxostat Clinical Study.....................................69 9.4 Clinical Efficacy in Real World Setting...................................................................70 9.5 Efficacy Conclusion.................................................................................................72 10.0 BENEFITS AND RISK ASSESSMENT.......................................................................73 10.1 Analysis of Condition..............................................................................................73 10.2 Current Treatment Options.......................................................................................73 10.3 Benefit.....................................................................................................................74 10.4 Risk and Risk Management......................................................................................75 10.4.1 Risk...................................................................................................................75 10.4.2 Risk Management..............................................................................................76 11.0 REFERENCES..............................................................................................................77 LIST OF IN-TEXT TABLES Table 1.a Adjudicated Major CV Events From Pooled Analysis of Phase 3 Randomized Controlled Studies (APEX, FACT, CONFIRMS).................................................4 Table 1.b Analysis of MACE and Each Event in the MACE Composite (mITT)..................5 Table 2.a Adjudicated Major CV Events From Pooled Analysis of Phase 3 Randomized Controlled Studies (APEX, FACT, CONFIRMS)...............................................15 Advisory Committee Briefing Materials: Available for Public Release Uloric (Febuxostat) Page 10of 82 AdvisoryCommittee Briefing Document 11 January 2019 Table 2.b Key US-FDA Interactions Related to the CVOT................................................16 Table 3.a Rates (PY) of Nonfatal MI, CV Death, and All Cause Deaths in Gout and No Gout...................................................................................................................18 Table 4.a Final Titrated Treatment Dose Overall and by Renal Function...........................30 Table 4.b CARES Subjects Study Drug Disposition..........................................................30 Table 4.c CARES Subjects Study Visit Disposition...........................................................31 Table 4.d Adjudicated MACE Composite: Primary Analysis (75% Interim) and Final Analysis of Primary Endpoint (mITT)................................................................33 Table 4.e Summary of Sensitivity Analysis for the Primary Endpoint (mITT)...................34 Table 4.f APTC Composite (CV death, Nonfatal MI, and Nonfatal Stroke): Analysis of Secondary Endpoint (mITT)...............................................................................37 Table 4.g Analysis of MACE and Each Event in the MACE Composite (mITT)................38 Table 4.h Summary of CV Deaths (mITT).........................................................................39 Table 4.i Adjudicated Causes of CV Death (mITT)...........................................................40 Table 4.j Summary of All-Cause Deaths (mITT)...............................................................43 Table 4.k Summary of All-Cause Mortality Incorporating Vital Status (mITT)..................44 Table 4.l CV Deaths by Baseline Renal Function..............................................................45 Table 4.m Characteristics of Subjects With Moderate Renal Impairment According to Dose Titration Requirements..............................................................................47 Table 4.n Comparison of Sensitivity Analyses With mITT Analysis for MACE and CV Death.................................................................................................................50 Table 4.o Analysis of MACE Using Multiple Imputation..................................................51 Table 5.a Thorough QT Study of Febuxostat at Maximal Therapeutic and Supratherapeutic Doses in Healthy Volunteers...................................................54 Table 5.b APTC Events in APEX and FACT.....................................................................55 Table 5.c Adjudicated APTC Events in CONFIRMS.........................................................56 Table 5.d Adjudicated Major CV Events Pooled Analysis of Phase 3 Randomized Controlled Studies (APEX, FACT, CONFIRMS)...............................................56 Table 9.a Proportion of Subjects With Serum Urate Levels <6.0 mg/dL and <5.0 mg/dL by Visit..............................................................................................................68 Table 9.b Proportion of Subjects With Gout Flares Requiring Treatment...........................68 LIST OF IN-TEXT FIGURES Figure 1.a Febuxostat Clinical Development Program Supporting Approval.........................3 Figure 3.a Serious Comorbidities Associated With Gout: NHANES 2007-2008.................18 Advisory Committee Briefing Materials: Available for Public Release