Efficacy and Safety of Evolocumab AMG 145 Monotherapy

Efficacy and Safety of Evolocumab (AMG 145) Monotherapy Compared With Ezetimibe and Placebo in Hypercholesterolemic Subjects: A Phase 3 Randomized Clinical Trial Michael J Koren, 1 Pernille Lundqvist, 2 Michael Bolognese, 3 Joel M Neutel, 4 Maria Laura Monsalvo, 5 Jingyuan Yang, 5 Jae B Kim, 5 Rob Scott, 5 Scott M Wasserman, 5 Harold Bays 6 for the MENDEL-2 Investigators 1 Jacksonville Center for Clinical Research, Jacksonville, FL, USA; 2 Center for Clinical and Basic Research, Ballerup, Denmark; 3 Bethesda Health Research Center, Bethesda, MD, USA; 4 Orange County Research Center, Tustin, CA, USA; 5 Amgen Inc. , Thousand Oaks, CA, USA; 6 L-MARC Research Center, Louisville, KY, USA March 29, 2014, Featured Clinical Research Session 400 American College of Cardiology, Washington DC

Background Ø Evolocumab, a fully human monoclonal antibody against PCSK 9, has emerged as a novel therapeutic option for lowering LDL-C in patients with hypercholesterolemia. Ø In Phase 2 studies, treatment with evolocumab was well tolerated and significantly reduced LDL-C in diverse groups of subjects, with LDL-C reductions maintained over 52 weeks. 1 -5 Ø Because statins up-regulate PCSK 9 production, evaluation of evolocumab as monotherapy is required to fully understand its pharmacodynamics and safety profile in populations not confounded by statin use or a history of statin intolerance, particularly at doses anticipated for use in clinical practice. 1. Lancet. 2012; 380: 1995 -2006. 2. Lancet. 2012; 380: 2007 -2017. 3. JAMA. 2012; 308: 2497 -2506. 4. Circulation. 2012; 126: 2408 -2417. 5. Circulation. 2014; 129: 234 -243. 2

The MENDEL-2 Study Ø Monoclonal Antibody Against PCSK 9 to Reduce Elevated LDL-C in Patients Currently Not Receiving Drug Therapy For Easing Lipid Levels-2 (NCT 01763827) Ø Design: A 12 week randomized, double-blind, placebo- and ezetimibe-controlled multicenter phase III study Ø Objective: To evaluate efficacy and safety of evolocumab in primary hypercholesterolemic patients not taking statins 3

MENDEL and MENDEL-2 Compared Characteristics Phase N Population Fasting LDL-C Range Background Lipid Therapy MENDEL-2 2 3 406 614 Framingham Risk < 10% ≥ 100 and < 190 mg/d. L None Evolocumab Dose • 70, 105, or 140 mg biweekly • 280, 350, or 420 mg monthly • 140 mg biweekly • 420 mg monthly Comparators • Placebo biweekly or monthly • Open ezetimibe 10 mg/day • Placebo biweekly or monthly • Blinded ezetimibe 10 mg/day Treatment Duration Evolocumab Delivery 12 wks 70 mg/ml vial and syringe in clinics 140 mg/ml prefilled autoinjector in clinic and home 4

MENDEL-2: Study Design 140 mg Evolocumab Q 2 W / Placebo PO QD N = 153 End of Study Screening Period with Placebo Injection Randomization 420 mg Evolocumab QM / Placebo PO QD N = 153 Placebo SC Q 2 W / 10 mg Ezetimibe PO QD N = 77 2: 2: 1: 1 Placebo SC QM / 10 mg Ezetimibe PO QD N = 77 Placebo SC Q 2 W / Placebo PO QD N = 77 Placebo SC QM / Placebo PO QD N = 78* Day 1 Week 2 Week 4 Week 6 Week 8 Week 10 Week 12 Week 14† Biweekly SC administration: Monthly SC administration: *1 patient was randomized but not dosed. †Phone call for AEs, SAEs. AEs, adverse events. EOS, end of study; QD, daily; Q 2 W, biweekly; QM, monthly. QM EOS Q 2 W EOS 5

MENDEL-2: Primary Endpoints Ø Co-primary endpoints: Percent change from baseline in LDL-C at week 12 and mean of weeks 10 and 12 Ø Secondary endpoints: At mean of weeks 10 and 12 and at week 12: § Percent change from baseline in Apo. B, Apo. A-I, lipoprotein(a), TG, and HDL-C § Percent of patients with LDL-C <70 mg/d. L Ø Key safety endpoints: § Treatment-emergent and serious adverse events § Muscle and hepatic enzyme elevations § Anti-evolocumab antibodies 6

MENDEL-2: Baseline Demographics Biweekly Monthly PBO Q 2 W + PBO QD (N = 76) PBO Q 2 W + EZE QD (N = 77) Evolocumab 140 mg Q 2 W + PBO QD (N = 153) 54 (10) 54 (11) 53 (14) 53 (11) 53 (13) 53 (12) Female, % 63 69 68 60 68 66 Race, white, % 83 82 86 81 78 84 NCEP risk categories*, % High Moderately high Moderate Low 0 8 33 59 0 4 47 49 1 3 41 55 3 5 26 67 1 5 40 53 1 5 34 60 Age (years), mean (SD) PBO QM + PBO QD (N = 78) PBO QM + EZE QD (N = 77) Evolocumab 420 mg QM + PBO QD (N = 153) *Risk category definitions: high, diagnosed CHD or risk equivalent; moderately high, 2 or more risk factors and Framingham risk score 10%-20%; moderate, 2 or more risk factors and Framingham risk score <10%; lower, 0 or 1 risk factor. CV, cardiovascular; EZE, ezetimibe; PBO, placebo; Q 2 W, biweekly; QM, monthly; QD, daily. 7

MENDEL-2: Baseline Lipids Biweekly Monthly PBO Q 2 W + PBO QD (N = 76) PBO Q 2 W + EZE QD (N = 77) Evolocumab 140 mg Q 2 W + PBO QD (N = 153) LDL-C*, mg/d. L, mean (SD) 140 (21) 143 (24) 142 (22) 144 (24) 144 (23) Apo. B, md/d. L, mean (SD) 104 (17) 107 (20) 105 (17) 107 (20) 106 (18) 108 (18) 21 (9, 49) 28 (11, 120) 20 (7, 58) 22 (7, 62) 28 (12, 64) 28 (9, 104) TG, mg/d. L, median (Q 1, Q 3) 114 (83, 178) 113 (84, 158) 112 (82, 148) 118 (86, 179) 117 (90, 159) 119 (83, 169) PCSK 9, ng/m. L, mean (SD) 281 (89) 270 (94) 272 (81) 270 (82) 265 (94) 274 (84) Lp(a), nmol/L, median (Q 1, Q 3) PBO QM + PBO QD (N = 78) PBO QM + EZE QD (N = 77) Evolocumab 420 mg QM + PBO QD (N = 153) *Determined by the Friedewald formula with reflexive testing via preparative ultracentrifugation when calculated LDL-C was <40 mg/d. L (1. 0 mmol/L) or triglyceride levels were >400 mg/d. L (3. 9 mmol/L). CV, cardiovascular; EZE, ezetimibe; PBO, placebo; Q 2 W, biweekly; QM, monthly; QD, daily. 8

MENDEL-2: Evolocumab Primary Endpoint Biweekly Dose Treatment Difference vs Placebo Mean Percent Change in LDL-C from Baseline Treatment Difference vs Ezetimibe Average at weeks 10 and 12 At week 12 -57% -39% P<0. 001 10 0 0. 1% – 10 – 18% – 20 – 30 – 40 – 57% – 60 Biweekly SC administration BL Day 1 Week 2 Week 4 Week 6 Week 8 Week 10 Week 12 Study Week Placebo (N = 76) Ezetimibe (N = 77) Evolocumab biweekly (N = 153) BL, baseline. Vertical lines represent the standard error around the mean. Plot is based on observed data with no imputation for missing values. P values are multiplicity adjusted. 9

MENDEL-2: Evolocumab Primary Endpoint Monthly Dose Treatment Difference vs Placebo Mean Percent Change in LDL-C from Baseline Treatment Difference vs Ezetimibe Average at weeks 10 and 12 At week 12 -57% -55% -40% -38% P<0. 001 10 – 1% 0 – 19% – 20 – 30 – 40 – 56% – 60 Monthly SC administration BL Day 1 Week 2 Week 4 Week 6 Week 8 Week 10 Week 12 Study Week Placebo (N = 78) Ezetimibe (N = 77) Evolocumab monthly (N = 153) BL, baseline. Vertical lines represent the standard error around the mean. Plot is based on observed data with no imputation for missing values. P values are multiplicity adjusted. 10

Percent Change in LDL-C (mg/d. L) Monthly Biweekly MENDEL-2: % Change in LDL-C from Baseline at Week 12 for Individual Patients Placebo (N = 76) Ezetimibe (N = 77) Evolocumab (N = 153) Placebo (N = 78) Ezetimibe (N = 77) Evolocumab (N = 153) 60 40 20 0 – 20 – 40 – 60 – 80 – 100 �Subjects with early termination of subcutaneous or oral study drug No notable difference in results for average at weeks 10 and 12 11

MENDEL-2: Other Lipids at Week 12 Apo. B Biweekly Monthly 0 Treatment difference (biweekly and monthly) vs placebo P < 0. 001 vs ezetimibe P < 0. 001 Treatment Difference, Mean (%) -10 -20 -30 -40 -50 -60 Treatment Difference, Median (%) Lp(a) – 33% – 34% – 48% Biweekly Monthly 0 -5 -10 Treatment difference (biweekly and monthly) vs placebo P < 0. 001 vs ezetimibe P < 0. 001 -15 Placebo Ezetimibe -20 -25 -30 – 20% – 18% – 16% Error bars represent standard error for mean treatment difference and 95% CI for median treatment difference. No notable difference in results for average at weeks 10 and 12. P values are multiplicity adjusted. 12

MENDEL-2: Other Lipids at Week 12 - II Treatment Difference, Median (%) Triglycerides 10 Biweekly Monthly Treatment difference (monthly) vs placebo P < 0. 001 (biweekly and monthly) vs ezetimibe P < 0. 05 0 -10 -20 – 6% – 8% – 13% Treatment Difference, Mean (%) Non-HDL-C – 18% -30 Biweekly Monthly 0 -10 Treatment difference (biweekly and monthly) vs placebo P < 0. 001 vs ezetimibe P < 0. 001 -20 -30 -40 – 33% – 35% Placebo Ezetimibe -50 -60 – 50% – 51% Error bars represent standard error for mean treatment difference and 95% CI for median treatment difference. No notable difference in results for average at weeks 10 and 12. P values are multiplicity adjusted. 13

MENDEL-2: Other Lipids at Week 12 - III 15 13 11 9 7 5 3 1 -1 -3 -5 -7 Treatment Difference, Median (%) HDL-C Apo. A-I Biweekly 9% 8% 6% 6% Biweekly Monthly Treatmemt Difference, Mean (%) 8 6 4 2 Monthly Treatment difference (monthly only) vs placebo P < 0. 01 vs ezetimibe P < 0. 05 5% 3% 3% Treatment difference (biweekly and monthly) vs placebo P < 0. 05 vs ezetimibe P < 0. 05 3% Placebo Ezetimibe 0 Error bars represent standard error for mean treatment difference and 95% CI for median treatment difference. No notable difference in results for average at weeks 10 and 12. P values are multiplicity adjusted. 14

MENDEL-2: Treatment Difference According to Subgroups at Week 12 Evolocumab Q 2 W vs Placebo vs Ezetimibe Evolocumab QM Vs Placebo Evolocumab QM vs Ezetimibe Sex Female Male Age <65 years ≥ 65 years Race White Black Other History of Met Syn Yes No Screening LDL-C <130 mg/d. L ≥ 130 mg/d. L Triglycerides <200 mg/d. L ≥ 200 mg/d. L Overall -100 -80 -60 -40 -20 0 Met Syn: metabolic syndrome -100 -80 -60 -40 -20 0 Percent Change from Baseline at Week 12 -100 -80 -60 -40 -20 0 15

MENDEL-2: Treatment Difference According to Subgroups at Week 12 Evolocumab Q 2 W vs Placebo vs Ezetimibe Evolocumab QM vs Placebo Evolocumab QM vs Ezetimibe Hypertension Yes No Smoker Yes No Baseline CHD risk factors <2 ≥ 2 Region North America Europe Other PCSK 9 <baseline median (261. 0 ng/m. L) ≥baseline median (261. 0 ng/m. L) Overall -100 -80 -60 -40 -20 0 Percent Change from Baseline at Week 12 -100 -80 -60 -40 -20 0 16

MENDEL-2: Safety and Tolerability Placebo (N = 154) Ezetimibe (N = 154) Evolocumab (N = 306) 68 (44) 70 (46) 134 (44) Common treatment-emergent AEs* Headache Diarrhea Nausea Urinary Tract Infection Constipation Nasopharyngitis Upper Respiratory Infection 4 (3) 6 (4) 1 (1) 2 (1) 4 (3) 3 (2) 4 (3) 5 (3) 3 (2) 1 (1) 6 (4) 5 (3) 10 (3) 9 (3) 8 (3) 7 (2) 6 (2) 5 (2) Serious AEs 1 (1) 4 (1) AEs leading to study drug discontinuation 6 (4) 5 (3) 7 (2) Deaths 0 (0) 8 (5) 7 (5) 16 (5) Muscle-related SMQ‡ Myalgia Musculoskeletal pain 6 (4) 3 (2) 2 (1) 5 (3) 3 (2) 1 (1) 8 (3) 3 (1) Neurocognitive AEs 0 (0) CK > 5 x ULN 2 (1) 0 (0) 2 (1) ALT or AST > 5 x ULN 2 (1) 0 (0) 1 (0. 3) NA NA 0 Adverse Events (AEs), n (%) Treatment-emergent AEs Potential injection site reactions † Anti-evolocumab antibodies § *Reported in ≥ 3% of patients in one or more treatment arms. †Reported using high-level term grouping, which includes injection site (IS) 17 rash, IS inflammation, IS pruritus, IS reaction, and IS urticaria. ‡Standard Med. DRA Queries. §Binding or neutralizing.

MENDEL-2: Conclusions Ø In the largest monotherapy trial with a PCSK 9 inhibitor to date, evolocumab biweekly (140 mg) or monthly (420 mg) rapidly and markedly lowered LDL-C over 12 weeks compared with placebo or ezetimibe in patients with hypercholesterolemia not taking statins. Ø Evolocumab 140 mg biweekly and 420 mg monthly dosing regimens are clinically equivalent. Ø Evolocumab treatment resulted in favorable changes in other lipoproteins. Ø Evolocumab was well tolerated over the 12 -week study § The overall incidence of treatment-emergent AEs, CK, and AST/ALT elevations was comparable across treatment groups. 18

MENDEL-2: Conclusions Ø Though we anticipate that evolocumab treatment will find use primarily as a treatment for high risk patients in conjunction with statins, MENDEL-2 demonstrated that evolocumab produces large LDL-C lowering effects as monotherapy. Ø These lipid effects occur consistently across several prespecified sub-groups. Ø The MENDEL-2 findings support future investigation of evolocumab in higher-risk patient population who might benefit from anti-PCSK 9 monotherapy. 19

Disclosures This study was funded by Amgen Inc. MJ Koren: Employee of Jacksonville Center for Clinical Research, which has received research grants from Amgen Inc. and other PCSK 9 -related research funding from Regeneron, Sanofi, Roche, and Pfizer. P Lundqvist: None. MA Bolognese: Research grant from Amgen Inc. JM Neutel: None ML Monsalvo, J Yang, JB Kim, R Scott, SM Wasserman: Employees of Amgen Inc. and own Amgen stock/stock options. H Bays: Research grant and consultant/advisory board from Amgen Inc. , and consultant/advisory board/speaker's bureau for many other pharmaceutical companies. Amgen Inc. provided editorial support for the development of this presentation. 20