Evaluation of Ranolazine in Patients with Type 2

Evaluation of Ranolazine in Patients with Type 2 Diabetes Mellitus and Chronic Stable Angina Results from the TERISA Randomized Clinical Trial Mikhail Kosiborod, Suzanne V. Arnold, John A. Spertus, Darren K. Mc. Guire, Yan Li, Patrick Yue, Ori Ben-Yehuda, Amos Katz, Phillip G. Jones, Ann Olmsted, Luiz Belardinelli, Bernard R. Chaitman On behalf of the TERISA Investigators

Funding • TERISA was sponsored by Gilead Sciences Inc. – Saint Luke’s Mid America Heart Institute received funding for independent data analysis from Gilead Sciences, Inc.

Background • Despite advances in therapy, chronic angina remains a common problem – Affects 8 million patients in the US – Negative impact on health status and Qo. L – Major driver of repeat hospitalizations and costs • Patients with diabetes have more extensive CAD, and greater angina burden than those without diabetes

Background • Ranolazine, an inhibitor of the late sodium current (late INa), is effective in treating chronic angina • Ranolazine may also lower fasting glucose and Hb. A 1 c • The anti-anginal efficacy of ranolazine in patients with diabetes has not been prospectively tested

TERISA: Primary Objective • Evaluate efficacy of ranolazine versus placebo on angina frequency in patients with type 2 diabetes, CAD, and chronic stable angina who remain symptomatic despite treatment with 1 or 2 antianginal medications

TERISA: Study Design • Run-in Phase: Single-blind placebo (4 weeks) • Treatment Phase: Randomized double-blind parallel group phase (8 weeks): ranolazine (target dose 1000 mg bid vs. matching placebo) Run-in Phase* Treatment Phase 4 weeks 8 weeks Wk 2 visit Wk 8 visit FU safety phone call Ranolazine Randomization Screening FU – 2 wks Wk 2 visit Wk 8 visit Placebo Antianginal Background Therapy *Subjects taking >2 antianginal medications or disallowed antianginal agents were allowed additional 2 week washout period prior to the run-in phase FU safety phone call

Study Endpoints • Primary: Average weekly number of angina episodes from weeks 2 -8 of treatment • Key Secondary: Average weekly number of SL NTG doses from weeks 2 -8 of treatment

Data Acquisition • Angina frequency and SL NTG use captured daily using electronic diary • Daily data transfer

TERISA Sites 105 Sites from 14 Countries

Enrollment and Randomization Assessed for Eligibility (n=1185) Excluded (n=236) • Not meeting inclusion criteria (n=43) • Failed run-in (n=193) Randomized (n=949) Randomized to Ranolazine (n=473) Randomized to Placebo (n=476) Discontinuation of Treatment (n=11) Analyzed (n=462) Analyzed (n=465)

Baseline Characteristics by Study Group Age (yr) Men (%) White (%) Hypertension (%) Dyslipidemia (%) Current smoking (%) Prior myocardial infarction (%) Prior angioplasty (%) Prior bypass graft surgery (%) Ranolazine n=462 Placebo n=465 63. 2 61. 3 98. 7 95. 0 79. 4 15. 4 75. 4 42. 7 18. 2 64. 2 61. 5 99. 4 95. 9 80. 3 16. 6 72. 7 38. 8 18. 9

Baseline Characteristics by Study Group Ranolazine n=462 Placebo n=465 Duration of diabetes (yr) 7. 2± 6. 7 7. 7± 7. 0 Hb. A 1 c (%) 7. 3± 1. 5 Glucose Lowering Medication (%) 93. 3 92. 7 Insulin (%) 17. 5 20. 6

Baseline Characteristics by Study Group Antianginal medications on 1 (%) on 2 (%) Beta blockers (%) Calcium channel blockers (%) Long acting nitrates (%) Statins (%) Antiplatelet agents (%) ACE-I/ARBs (%) Diary compliance - median % (IQR) Ranolazine n=462 Placebo n=465 56. 1 43. 9 90. 5 26. 8 34. 8 82. 5 89. 8 88. 1 98 (95 -98) 55. 7 44. 3 89. 9 30. 8 32. 5 82. 4 86. 5 87. 5 98 (95 -98)

Primary Endpoint Ranolazine n=462 Placebo n=465 p-value Least squares mean (95% CI) Angina frequency, baseline (#/wk) 6. 6 (6. 3 -7. 0) 6. 8 (6. 4 -7. 2) 0. 54 Angina frequency, on treatment (#/wk) 3. 8 (3. 6 -4. 1) 4. 3 (4. 0 -4. 5) 0. 008

Weekly Angina Frequency by Study Group Treatment Phase Weekly Angina Frequency Run In Phase 6 4 2 Placebo p=0. 008 Ranolazine 0 -2 0 2 Study Week 4 6 8

Key Secondary Endpoint Ranolazine n=462 Placebo n=465 p-value Least squares mean (95% CI) SL NTG doses, baseline – (#/wk) 4. 1 (3. 7 -4. 6) 4. 5 (4. 1 -5. 0) 0. 27 SL NTG doses, on treatment (#/wk) 1. 7 (1. 6 -1. 9) 2. 1 (1. 9 -2. 3) 0. 003

SL NTG Doses Weekly SL NTG Doses 4 Treatment Phase Run In Phase 3 2 1 0 Placebo p=0. 003 Ranolazine -2 0 2 Study Week 4 6 8

Subgroup Analyses of the Primary End Point of Weekly Angina Frequency Ranolazine better Placebo better Other Russia, Ukraine, Belarus p for interaction 0. 016 Pre-specified stratifications Other: 3. 1 vs. 4. 1 (ranolazine vs. placebo), p=0. 002 Russia, Ukraine, Belarus: 4. 1 vs. 4. 3 (ranolazine vs. placebo), p=0. 31 0. 7 0. 8 0. 9 1 1. 2 Incidence Density Ratio

Subgroup Analyses of the Primary End Point of Weekly Angina Frequency Ranolazine better Placebo better Other Russia, Ukraine, Belarus 2 antianginal medications 1 antianginal medications Prespecified ≥ 3 baseline episodes stratifications < 3 baseline episodes p for interaction 0. 016 0. 89 0. 85 Age ≥ 65 Age < 65 Men Women Prior PCI No Prior PCI 0. 97 0. 46 0. 61 Prior CABG No Prior CABG 0. 28 0. 7 0. 8 0. 9 1 1. 2 Incidence Density Ratio

Exploratory Analysis – Hb. A 1 c Ranolazine better Placebo better p for interaction Hb. A 1 c >6 Hb. A 1 c ≤ 6 0. 046 Hb. A 1 c >6. 5 Hb. A 1 c ≤ 6. 5 0. 047 Hb. A 1 c >7 Hb. A 1 c ≤ 7 0. 022 Hb. A 1 c >7. 5 Hb. A 1 c ≤ 7. 5 0. 041 Hb. A 1 c >8 Hb. A 1 c ≤ 8 0. 038 0. 7 0. 8 0. 9 1 1. 1 Incidence Density Ratio 1. 2

Safety and Tolerability Serious Adverse Events Serious adverse event Death Nonfatal myocardial infarction Stroke/transient ischemic attack Unstable angina or coronary revascularization Notable non-serious adverse events Dizziness Nausea Headache Constipation Hypoglycemia Any Adverse Event Ranolazine Placebo n=470 n=474 number (%) 16 (3. 4) 20 (4. 2) 3 (0. 6) 2 (0. 4) 1 (0. 2) 3 (0. 6) 1 (0. 2) 4 (0. 8) p-value 0. 51 0. 69 0. 62 0. 37 6 (1. 3) 7 (1. 5) 0. 79 17 (3. 6) 7 (1. 5) 8 (1. 7) 3 (0. 6) 126 (26. 8) 6 (1. 3) 2 (0. 4) 9 (1. 9) 2 (0. 4) 0 (0. 0) 105 (22. 2) 0. 019 <0. 001 0. 63 0. 063 0. 12 0. 096

Conclusions • Ranolazine was more effective than placebo in reducing angina frequency and SL NTG use in patients with type 2 diabetes, CAD and chronic angina • The therapeutic effectiveness of ranolazine was greater – In patients enrolled outside of Russia, Ukraine and Belarus – In those with higher baseline Hb. A 1 c • Future studies are needed to explore potential dual effects of ranolazine on angina and glucose control in patients with type 2 diabetes

JACC Online • Evaluation of Ranolazine in Patients with Type 2 Diabetes Mellitus and Chronic Stable Angina. Results from the TERISA randomized clinical trial • Journal of the American College of Cardiology (2013), doi: 10. 1016/j. jacc. 2013. 02. 011.

Back-up Slides

Weekly Angina Frequency by Study Group Stratified by Geographic Region Weekly Angina Frequency Run In Phase 6 4 2 0 Placebo Ranolazine Russia, Ukraine, Belarus (n=667) -2 Weekly Angina Frequency Treatment Phase 6 4. 1 vs. 4. 3 (ranolazine vs. placebo), p=0. 31 0 Run In Phase 2 Study Week 4 6 8 Treatment Phase 4 2 0 Placebo Ranolazine 3. 1 vs. 4. 1 (ranolazine vs. placebo), p=0. 002 Other (n=282) -2 0 2 Study Week 4 6 8

Secondary Endpoints Ranolazine n=462 Placebo n=465 p-value Least squares mean (95% CI) Percentage of angina-free days - % 67 (65 -70) 64 (61 -67) 0. 068 Subjects with ≥ 50% reduction in angina episodes - % 47 (43 -51) 42 (38 -46) 0. 034 SF-36 Mental Component Score, change from baseline 1. 0 (0. 2 -1. 8) 1. 1 (0. 3 -1. 9) 0. 77 SF-36 Physical Component Score, change from baseline 2. 9 (2. 3 -3. 5) 1. 9 (1. 3 -2. 5) 0. 005 PGIC scale score 4. 0 (3. 8 -4. 2) 3. 9 (3. 7 -4. 1) 0. 41

Baseline Characteristics of Patients by Geographic Region Russia/Ukraine/Belarus n=659 Other n=268 P Age (yr) 63. 3± 8. 5 64. 6± 8. 5 0. 039 Men (%) 58. 9% 67. 5% 0. 014 White (%) 100. 0% 96. 6% <0. 001 Hypertension (%) 96. 2% 93. 6% 0. 089 Dyslipidemia (%) 77. 8% 84. 9% 0. 017 Current smoking (%) 15. 5% 17. 2% 0. 525 Prior MI (%) 76. 4% 68. 3% 0. 011 Prior angioplasty (%) 30. 4% 66. 3% <0. 001 Prior CABG (%) 17. 5% 21. 3% 0. 175

Baseline Characteristics of Patients by Geographic Region Russia/Ukraine/Belarus n=659 Other n=268 P Duration of diabetes (yr) 6. 6± 6. 3 9. 5± 7. 6 <0. 001 Hb. A 1 c (%) 7. 3± 1. 5 0. 561 Glucose Lowering Medication (%) 93. 2% 92. 5% 0. 732 Insulin (%) 16. 2% 26. 1% <0. 001

Baseline Characteristics of Patients by Geographic Region Russia/Ukraine/Belarus n=659 Other n=268 Antianginal medications P 0. 002 on 1 (%) 59. 0% 48. 1% on 2 (%) 41. 0% 51. 9% Beta blockers—no. (%) 89. 2% 92. 5% 0. 125 Calcium channel blockers—no. (%) 28. 7% 29. 1% 0. 897 Long acting nitrates—no. (%) 34. 3% 32. 1% 0. 520 Statins—no. (%) 82. 5% 82. 1% 0. 868 Antiplatelet agents—no. (%) 88. 2% 88. 1% 0. 965 ACE-I/ARBs—no. (%) 89. 8% 82. 8% 0. 003 Weekly angina frequency, baseline 7. 2± 4. 5 6. 3± 3. 8 0. 007

Incidence Density Ratios for Angina Frequency Omitting Individual Countries IDR for overall cohort: 0. 894

Incidence Density Ratio, Angina Frequency Incidence Density Ratios for Angina Frequency Omitting Individual Sites (Russian Sites in Red) IDR for overall cohort: 0. 894 Site

Enrollment in TERISA by Country Belarus Bulgaria Canada Czech Republic Georgia Germany Israel Poland Russia Serbia Slovakia Slovenia Ukraine United States Total Number of Subjects 1 18 3 8 75 3 23 89 537 5 20 1 121 23 927 Number of subjects from each country who comprised the full analytic sample, from which the efficacy analyses were derived

The Cycle of Myocardial Ischemia: Ischemia Begets Ischemia ↓Oxygen Supply: Demand MVO 2 Supply Contracture ( LVEDP) Arrhythmias Ischemia Diabetic Heart (↑ Glucose) Ranolazine Hypothesis ↑ Late INa Na+i & Ca++i Overload Hypothesis : Glucose increases p. Ca. MKII, which increases late INa Modified from Belardinelli L. et al. Eur Heart J. Suppl. 2006

Diabetic Cardiomyopathy: Dual Benefit of Ranolazine High Glucose Late INa Hypothesis * Hypothesis : Glucose increases p. Ca. MKII, which increases late INa * Nishio et al JMCC 52 (2012) 1103– 1111 * Luo and Anderson et al JCI, 2013 * Mourouzis et al Unpublished data 2013