I Articles
`
`+~ Effects of eicosapentaenoic acid on major coronary events in
`hypercholesterolaemic patients (JELIS): a randomised open(cid:173)
`label, blinded endpoint analysis
`
`MitsuhiroYokoyama, Hideki Origasa, Masunori Matsuzaki, Yuji Matsuzawa, Yasushi Saito, Yuichi Ishikawa, Shinichi Oikawa,jun Sasaki,
`Hitoshi Hishida, Hiroshige ltakura, Toru Kita, Akira Kitabatake, Noriaki Nakaya, Toshiie Sakata, Kazuyuki Shimada, Kunio Shirato, for the
`japan EPA lipid intervention study (}ELIS) Investigators
`
`Summary
`Lancet 2007; 369: 1090-98 Background Epidemiological and clinical evidence suggests that an increased intake of long-chain n-3 fatty acids
`See Comment page 1062 protects against mortality from coronary artery disease. We aimed to test the hypothesis that long-term use of
`Kobe university, Kobe, Japan eicosapenlaenoic acid (EPA) is effective for prevention of major coronary events in hypercholesterolaemic patients in
`(M Yokoyama MD); Division of Japan who consume a large amount of fish.
`Clinical Epidemiology and
`Biostatistics, Toyama
`University, Toyama, Japan
`(H Origasa PhD); Yamaguchi
`University, Yamaguchi, Japan
`(M Matsuzaki MD); Sumitomo
`Hospital, Osaka, japan
`(Y Matsuzawa MD); Chiba
`University, Chiba, japan
`(Y Saito MD); Kobe University,
`Kobe, japan (Y Ishikawa MD);
`Nippon Medical School, Tokyo,
`Japan (S Oikawa MD);
`International University of
`Health and Welfare Graduate
`School of Public Health
`Medicine, Fukuoka, Japan
`(J Sasaki MD); Fujita Health
`University School of Medicine,
`Aichi, japan (H Hishida MD);
`lbaraki Christian University,
`lbaraki,japan (H ltakura MD);
`Kyoto University, Kyoto, Japan
`(T Kita MD); Showa Hospital,
`Hyogo, Japan
`(A Kitabatake MD); Nakaya
`Clinic, Tokyo, japan
`(N Nakaya MD); Nakamura
`Gakuen University, Fukuoka,
`Japan (T Sakata MD);Jichi
`Medical School, Tochigi, Japan
`(K Shimada MD); and Saito
`Hospital, Miyagi, Japan
`(K Shirato MD)
`
`Methods 18 645 patients with a total cholesterol of 6 · 5 mmol/L or greater were recruited from local physicians
`throughout Japan between 1996 and 1999. Patients were randomly assigned to receive either 1800 mg of EPA daily
`with statin (EPA group; n=9326) or statin only (controls; n=9319) with a 5-year follow-up. The primary endpoint was
`any major coronary event, including sudden cardiac death, fatal and non-fatal myocardial infarction, and other non(cid:173)
`fatal events including unstable angina pectoris, angioplasty, stenting, or coronary artery bypass grafting. Analysis was
`by intention-to-treat. The study was registered at clinicaltrials.gov, number NCT00231738.
`
`Findings At mean follow-up of 4 · 6 years, we detected the primary endpoint in 262 (2 · 8%) patients in the EPA group
`and 324 (3 · 5%) in controls-a 19% relative reduction in major coronary events (p=O · 011). Post-treatment LDL
`cholesterol concentrations decreased 25%, from 4.7 mmol/L in both groups. Serum LDL cholesterol was not a
`significant factor in a reduction of risk for major coronary events. Unstable angina and non-fatal coronary events were
`also significantly reduced in the EPA group. Sudden cardiac death and coronary death did not differ between groups.
`In patients with a history of coronary artery disease who were given EPA treatment, major coronary events were
`reduced by 19% (secondary prevention subgroup: 158 [8-7%] in the EPA group vs 197 [10-7%] in the control group;
`p=O · 048). In patients with no history of coronary artery disease, EPA treatment reduced major coronary events by
`18%, but this finding was not significant (104 [1·4%] in the EPA group vs 127 [1· 7%] in the control group; p=0· 132).
`
`Interpretation EPA is a promising treatment for prevention of major coronary events, and especially non-fatal coronary
`events, in Japanese hypercholesterolaemic patients.
`
`Correspondence to:
`Dr Mitsuhiro Yokoyama,
`Cardiovascular Medicine
`Division, Department of Internal
`Medicine, Kobe University
`Graduate School of Medicine,
`7-5-2, Kusunoki-cho, Chuo-ku,
`Kobe, 650-0017 Japan
`yokoyama@med.kobe-u.ac.jp
`
`Introduction
`Epidemiological and clinical evidence suggests a
`inverse association between
`long-term
`significant
`intake of long-chain n-3 polyunsaturated fatty acids,
`especially
`eicosapentaenoic
`acid
`(EPA)
`and
`docosahexaenoic acid (DHA), and mortality associated
`with coronary artery disease.1-7 Thus, the consumption
`of fish or fish-oil could protect against major events
`associated with coronary artery disease, especially fatal
`myocardial infarction and sudden cardiac death. Two
`large-scale secondary prevention trials, the Diet and
`Reinfarction Trial and the Gruppo Italiano per lo Studio
`della Sopravivenza nell' Infarto Miocardico-Prevenzione
`Trial, reported that increased consumption of fish or
`fish-oil
`supplements
`reduced coronary death
`in
`postinfarction patients. 8
`9 No randomised trials have
`·
`examined the effects of n-3 polyunsaturated fatty acids
`on major coronary events in a high-risk, primary
`prevention population.
`EPA ethyl ester, which is purified from n-3 poly(cid:173)
`unsaturated fatty acids present in fish oil, is approved
`
`by Japan's Ministry of Health, Labour, and Welfare as a
`treatment for hyperlipidaemia and peripheral artery
`include
`disease. The biological functions of EPA
`reduction of platelet aggregation, 10
`11 vasodilation,12.13
`·
`anti proliferation, 14 plaque-stabilisation, 15 and reduction
`in lipid action. 1617 Therefore the preventive effects of
`EPA on major cardiovascular events are of both clinical
`interest and therapeutic importance.
`Primary and secondary prevention trials have proved
`that cholesterol-lowering treatment with inhibitors of
`3-hydroxy-3-methylglutaryl coenzyme A (HMG CoA)
`reductase-statins-reduces
`the
`risk of all-cause
`mortality and major cardiovascular events in patients
`with a wide range of cholesterol concentrations, whether
`or not they have had coronary artery disease.18-
`21 Thus,
`statins are now established as the first-line treatment
`for hyperlipidaemia. 22 Preliminary data for treatment
`with a combination of n-3 polyunsaturated fatty acids
`and statins have shown beneficial effects on the lipid
`profiles of patients with a mixed type of hyper(cid:173)
`lipidaemia;23-25 however, no major long-term inter-
`
`1090
`
`www.thelancet.com Vol 369 March 31, 2007
`
`ICOSAPENT DFNDTS00007157
`
`Hikma Pharmaceuticals
`
`IPR2022-00215
`
`Ex. 1005, p. 1 of 9
`
`

`

`I
`
`Articles
`
`ventional trial has yet investigated whether the addition
`of EPA to conventional statin treatment would yield an
`incremental clinical benefit. The Japan EPA Lipid
`Intervention Study (JELIS) tests the hypothesis that
`long-term use of EPA is effective in reduction of major
`coronary events in Japanese hypercholesterolaemic
`patients given statins.
`
`Methods
`Study design and patients
`We did a prospective, randomised open-label, blinded
`endpoint evaluation (PROBE). 26 Our study design, and
`inclusion and exclusion criteria are described in detail
`elsewhere. 27 We recruited 19466 hypercholesterolaemic
`patients through local physicians from all regions of
`Japan between November, 1996, and November, 1999.
`Figure 1 shows the trial profile. The participants
`consisted of 5859 men
`(aged 40-75 years) and
`12786 postmenopausal women (aged up to 75 years),
`with or without coronary artery disease, which was
`defined as previous myocardial infarction, coronary
`interventions, or confirmed angina pectoris. Informed
`written consent was obtained from all eligible patients
`before random assignment to either the EPA treatment
`or control groups.
`Eligibility criteria were total cholesterol concentration of
`6 · 5 mmol/L or greater, which corresponded to a LDL
`cholesterol of 4·4 mmol/L or greater. Exclusion criteria
`were: acute myocardial infarction within the past 6 months,
`unstable angina pectoris, a history or complication of
`serious heart disease (such as severe arrhythmia, heart
`failure, cardiomyopathy, valvular disease, or congenital
`disease), cardiovascular reconstruction within the past
`6 months, cerebrovascular disorders within the past
`6 months, complications of serious hepatic or renal
`disease, malignant disease, uncontrollable diabetes,
`hyperlipidaemia due to other disorders, hyperlipidaemia
`caused by drugs such as steroid hormones, haemorrhage
`(including haemophilia, capillary fragility, gastrointestinal
`ulcer, urinary tract haemorrhage, haemoptysis, and
`vitreous haemorrhage), haemorrhagic diathesis, hyper(cid:173)
`sensitivity to the study drug formulation, patients' intention
`to undergo surgery, and judgment by the physician in
`charge that a patient was inappropriate for the study.
`The primary endpoint was any major coronary event,
`including sudden cardiac death, fatal and non-fatal
`myocardial
`infarction, and other non-fatal events
`including unstable angina pectoris,
`angioplasty,
`stenting, or coronary artery bypass grafting. Secondary
`endpoints (all-cause mortality, mortality and morbidity
`of coronary artery disease, stroke, peripheral artery
`disease, and cancer) are not reported here.
`
`Procedures
`We used the statistical coordination centre at the Toyama
`Medical and Pharmaceutical University to manage
`patient registration (including confirmation of eligibility
`
`19466 patients
`
`18 645 randomly assigned
`
`821 did not meet inclusion
`criteria
`805 refused to give consent
`16 other reasons
`
`9319 controls
`
`9326 EPA treatment
`
`791 discontinued observation
`98 losttofollow-up
`273 withdrew consent
`420 other reasons
`
`883 discontinued observation
`99 losttofollow-up
`361 withdrew consent
`423 other reasons
`
`9319 analysed by intention(cid:173)
`to-treat
`
`9326 analysed by intention(cid:173)
`to-treat
`
`Figure 1:Trial profile
`
`criteria), operation of the randomisation scheme, and
`data management. We used permuted-block random(cid:173)
`isation with a block size of four. Blocks were assigned
`according to the number of participants enrolled at each
`centre. Patients were divided into two subgroups: one
`with coronary artery disease (secondary prevention;
`n=3664) and one without (primary prevention; n=14981),
`and stratified accordingly. Patients were randomly
`assigned to receive EPA with statin (EPA group) or statin
`alone (controls). All patients first underwent 4-8 weeks
`of washout from antihyperlipidaemic drugs. Patients also
`received appropriate dietary advice.
`All patients received 10 mg of pravastatin or 5 mg of
`simvastatin once daily as first-line treatment. These
`statins were available in Japan at the initiation of this
`study, and these doses were recommended by the
`Ministry of Health, Labour, and Welfare. For serious
`hypercholesterolaemia (defined as uncontrolled), this
`daily dose was increased to 20 mg pravastatin or 10 mg
`simvastatin. No treatment with other antihyperlipidaemic
`drugs was allowed during the study. EPA was given at a
`dose of 600 mg, three times a day after meals (to a total of
`1800 mg per day). We used capsules that contained
`300 mg of highly purified (>98%) EPA ethyl ester
`(Machida Pharmaceuticals, Tokyo, Japan).
`Local physicians monitored compliance with dietary
`advice and medication, and noted adverse events at
`every clinic visit. Clinical endpoints and severe adverse
`events reported by local physicians were checked by
`members of a regional organising committee in a
`blinded fashion. Then, an endpoints adjudication
`committee (see webappendix), consisting of three SeeOnlineforwebappend1x
`expert cardiologists and one expert neurologist,
`confirmed them once a year without knowledge of the
`
`www.thelancet.com Vol 369 March 31, 2007
`
`1091
`
`ICOSAPENT DFNDTS00007158
`
`Hikma Pharmaceuticals
`
`IPR2022-00215
`
`Ex. 1005, p. 2 of 9
`
`

`

`I
`
`Articles
`
`Age (years)
`
`Male
`
`BMI (kg/m')
`
`Controls (n=9319)
`
`EPA treatment
`(n=9326)
`
`61(9)
`
`61 (8)
`
`2908 (31%)
`
`2951 (32%)
`
`24 (3)
`
`24(3)
`
`Cardiovascular history
`
`Myocardial infarction
`
`502 (5%)
`
`Angina
`
`CABGor PTCA
`
`Risk factors
`
`Smoking
`
`Diabetes
`
`Hypertension
`
`Serum lipid values
`
`1484(16%)
`
`433 (5%)
`
`1700(18%)
`
`1524(16%)
`
`3282 (35%)
`
`548 (6%)
`
`1419 (15%)
`
`462 (5%)
`
`1830(20%)
`
`1516 (16%)
`
`3329 (36%)
`
`Total cholesterol (mmol/L)
`
`7·11 (0·68)
`
`LDL-cholesterol (mmol/L)
`
`4-70 (0·75)
`
`HDL-cholesterol (mmol/L)
`
`1-51 (0-44)
`
`7·11 (0·67)
`
`4·69 (0·76)
`
`1·52 (0·46)
`
`Triglyceride (mmol/L)'
`
`1-74 (1·25-2-49)
`
`1·73 (1·23-2·48)
`
`135 (21)
`
`79(13)
`
`135 (21)
`
`79 (13)
`
`Blood pressure
`
`Systolic (mm Hg)
`
`Diastolic(mm Hg)
`
`HMGCoARI
`
`Pravastatin
`
`Si mvastati n
`
`Other statin
`
`Medication use
`
`Antiplateletagent
`
`Calcium antagonist
`B blocker
`Other anti hypertensive
`agents
`
`5553 (60%)
`
`3417(37%)
`
`128 (1%)
`
`1342 (14%)
`
`2837(30%)
`
`791 (8%)
`
`2424(26%)
`
`5523 (60%)
`
`3272 (36%)
`
`110 (1%)
`
`1258 (13%)
`
`2796 (30%)
`
`794(9%)
`
`2366 (25%)
`
`863 (9%)
`
`1081 (12%)
`
`Nitrate
`
`926 (10%)
`
`Hypoglycaemic agents
`
`1126 (12%)
`
`SeeOnlineforwebtable
`
`Data are number of patients(%) or mean (SD), unless otherwise indicated.
`CABG=coronary-artery bypass grafting. PTCA=percutaneous transluminal
`coronary angioplasty. LDL=low-density lipoprotein. HDL=high-density
`lipoprotein. IQR=interquartile range. HMG CoA Rl=3-hydroxy-3-methylglutaryl
`coenzyme A reductase in hi bit or. BMI "'body-mass index, which is weight in kg
`divided by the square of height in metres. 'Median (IQR).
`
`Table 1: Baseline characteristics
`
`treatment allocation. The study was approved by an
`external data and
`safety monitoring board, by
`institutional review boards at all hospitals, and by
`regional organising committees. The data and safety
`monitoring board also monitored the rate of endpoints
`twice during the study, in March, 2002, and March,
`2004. The study was followed up until November, 2004,
`because both interim analyses did not reach the
`stopping boundary.
`We sampled blood to measure serum lipid at 6 and
`12 months, and then every year until the final follow-up
`visits. Plasma total fatty acid concentrations for all
`patients who gave informed consent were measured
`with capillary gas chromatography every year at a central
`laboratory.
`
`Statistical analysis
`We used a two-sided test at the 5% significance level to
`estimate that the number of enrolled patients would
`give the study a statistical power of 80% for detection of
`a relative reduction of25% in the primary endpoint rate,
`when the EPA group was compared with controls. The
`event rate of the primary endpoint in the control group
`was assumed to be 0 · 58% per year for primary
`prevention and 2 · 13% per year for secondary prevention;
`the proportion of primary and secondary prevention
`strata was assumed to be 4:1. The accrual period was
`assumed to be 3 years with a follow-up of 5 years for all
`patients. All analyses were based on the intention-to(cid:173)
`treat principle. Time-to-event data were analysed with
`the Kaplan-Meier method and the log-rank test. The
`hazard ratio and its 95% confidence interval were
`computed with the Cox proportional hazard model. We
`did subgroup analyses with a model that included an
`interaction
`term corresponding
`to
`the
`test
`for
`heterogeneity in effects. Changes in lipid values were
`compared by repeated measures of ANOVA. Data were
`analysed with SAS statistical software (version 8.12).
`
`Role of the funding source
`The sponsor had no role in the study design, data
`collection, data analysis, data interpretation, or writing of
`the report. The JELIS steering committee had full access
`to all the data in the study and had final responsibility for
`the decision to submit for publication.
`
`Results
`Patients were monitored for an average of 4 · 6 years
`(SD 1·1). Table 1 shows baseline characteristics of the
`treatment groups. The mean age of all patients was
`61 years and 12 786 patients (69%) were women. Mean
`concentrations of total cholesterol and triglyceride were
`7 · l rnrnol/L and J. 7 rnrnol/L; and mean LDL and HDL
`cholesterol concentrations were 4· 7 mmol/L and
`1·5 mmol/L, respectively. The webtable shows baseline
`characteristics for primary and secondary prevention
`subgroups. Of 3664 patients with documented coronary
`artery disease, 1050 had a history of myocardial
`infarction, 2903 of angina pectoris, and 895 angioplasty,
`stenting, or coronary artery bypass grafting.
`Average doses were pravastatin 10 · 0 mg daily (SD
`9·1) and simvastatin 5·6 mg daily (1·8). 16449 (90%)
`patients took 10 mg pravastatin or 5 mg simvastatin.
`The 5-year follow-up rate was 16 971 (91%). Similar
`proportions of participants remained compliant in each
`treatment group. Study drug regimens were maintained
`until trial termination by 6151 (73%) of controls and in
`the treatment group 5883 (71%) of patients continued
`to take EPA and 6136 (74%) continued to take statin.
`586 patients (262 assigned to EPA and 324 controls)
`reached the composite primary endpoint. Figure 2
`shows Kaplan-Meier curves for the primary endpoint.
`The 5-year cumulative rate of major coronary events
`
`1092
`
`www.thelancet.com Vol 369 March 31, 2007
`
`ICOSAPENT DFNDTS00007159
`
`Hikma Pharmaceuticals
`
`IPR2022-00215
`
`Ex. 1005, p. 3 of 9
`
`

`

`Articles I
`
`Hazard ratio: 0·81 (0·657-0·998)
`p=0·048
`
`4
`
`Years
`
`A
`
`4
`
`...... Control
`-
`EPA
`
`B
`
`Hazard ratio: 0·81 (0·69-0·95)
`p=O·Oll
`
`4
`
`Years
`
`c
`
`12.0
`
`8·0
`
`4·0
`
`/''
`,vf
`:··· ,/·
`... r>~_..-/ Hazard ratio: 0-82 (0-63-1.06)
`
`,/'' ,...
`
`p=0·132
`
`o+'-~~~~~~~~~~~~~
`4
`0
`
`0
`
`0
`
`Years
`
`Numbers at risk
`
`Control group
`Treatment group
`
`9319
`9326
`
`8931
`8929
`
`8671
`8658
`
`8433
`8389
`
`8192
`8153
`
`7958
`7924
`
`7478
`7503
`
`7204
`7210
`
`7103
`7020
`
`6841
`6823
`
`6678
`6649
`
`6508
`6482
`
`1841
`1823
`
`1727
`1719
`
`1658
`1638
`
`1592
`1566
`
`1514
`1504
`
`1450
`1442
`
`Figure 2: Kaplan-Meier estimates of incidence of coronary events inthe total study population (panel A), the primary prevention arm (panel B) andthe secondary prevention arm (panel C)
`
`was 2-8% in the EPA group and 3-5% in controls,
`resulting in a significant relative risk reduction of 19%
`in the EPA group (p=O · 011). Figure 3 shows that EPA
`treatment was associated with a significant reduction of
`24% in the frequency of unstable angina. The occurrence
`of coronary death or myocardial infarction was not
`significantly lower (22%) in the EPA group than in
`controls. The frequency of fatal or non-fatal myocardial
`infarction was not significantly reduced (23%) in the
`EPA group; however, that of non-fatal coronary events
`(including non-fatal myocardial infarction, unstable
`angina, and events of angioplasty, stenting, or coronary
`artery bypass grafting) was significantly lower (19%) in
`the EPA group than in controls.
`Table 2 sets out major coronary events in the two
`treatment groups
`for comparison with
`specific
`background characteristics of all populations. For
`example, we grouped patients according to their LDL
`cholesterol at baseline. The relative reduction in major
`coronary events risk in the EPA group was of a similar
`magnitude in patients with different ranges of LDL
`cholesterol values, suggesting that LDL cholesterol is
`not an important factor in reduction of risk for major
`coronary events.
`In the primary prevention subgroup, EPA treatment
`was associated with a non-significant 18% reduction in
`major coronary events. Figure 3 shows the non(cid:173)
`significant reductions of18%, 21%, and 20% in coronary
`death or non-fatal myocardial infarction, fatal or non(cid:173)
`fatal myocardial infarction, and non-fatal coronary
`events,
`respectively.
`In the secondary prevention
`subgroup, allocation
`to
`the EPA
`treatment was
`associated with a significant 19% reduction in major
`coronary events. EPA
`treatment was also associated
`with a significant 28% reduction in the incidence of
`unstable angina. This treatment also produced non(cid:173)
`significant reductions of25%, 25%, and 18% in coronary
`
`infarction, fatal or non-fatal
`death or myocardial
`myocardial infarction, and non-fatal coronary events,
`respectively.
`In the other analyses, stroke occurred in 162 (1·7%)
`controls and 166 (1·8%) patients given EPA. Figure 3
`shows that the frequency ofischaemic and haemorrhagic
`strokes did not differ between the two treatment groups,
`and neither did all-cause mortality.
`Figure 4 summarises the change in lipid values after
`treatment. Total and LDL cholesterol at the last clinic
`visit decreased significantly by 19% and 25% from
`baseline in both groups, respectively. Triglyceride
`decreased significantly by 9% from baseline in the EPA
`group and by 4% in controls (p<O · 0001 between
`groups). Both treatments produced only small changes
`in HDL cholesterol. The fatty acid concentrations at
`baseline were the average values for all patients who
`gave informed consent in the control group (n=8076)
`and the EPA group (n=8321). Plasma EPA at baseline
`was 2 · 9% of total molecules of fatty acids (mol %). To
`assess the effect of EPA treatment, plasma fatty acid
`values were compared for all patients who were still
`compliant after 5 years of observation (controls: n=4854,
`EPA group: n=4970). Plasma EPA concentration and
`the ratio of EPA to arachidonic acid at baseline were
`93 mg/Land 0 · 60 in controls, and 97 mg/Land 0 · 63 in
`the EPA group, respectively. Plasma EPA concentration
`and the ratio of EPA to arachidonic acid at year 5 were
`93 mg/L and 0 · 59 in controls. On the other hand,
`plasma EPA concentration at year 5 was 169 mg/L in
`the EPA group, which was a 70% increase from baseline.
`The ratio of EPA to arachidonic acid increased two-fold
`from 0 · 63 to 1·23 in the EPA group. Similar results
`were reported previously. 11
`8
`"
`Table 3 shows that a quarter of patients in the EPA
`group had adverse experiences related to treatment,
`compared with about a fifth of controls. Rates of
`
`www.thelancet.com Vol 369 March 31, 2007
`
`1093
`
`ICOSAPENT DFNDTS00007160
`
`Hikma Pharmaceuticals
`
`IPR2022-00215
`
`Ex. 1005, p. 4 of 9
`
`

`

`I Articles
`
`Event
`
`All patients
`Major coronary events
`
`Items of account
`
`Sudden cardiac death
`Fatal Ml
`Non-fatal Ml
`Unstable angina
`CABGorPTCA
`Combined endpoint
`Coronary death or Ml
`Fatal Ml or nonfatal Ml
`
`Coronary death
`Non-fatal coronary events
`
`Primary prevention of CAD
`Major coronary events
`
`Items of account
`
`Sudden cardiac death
`Fatal Ml
`Non-fatal Ml
`Unstable angina
`CABGorPTCA
`Combined endpoint
`Coronary death or Ml
`Fatal Ml or nonfatal Ml
`
`Coronary death
`Non-fatal coronary events
`
`Secondary prevention of CAD
`
`Major coronary events
`Items of account
`Sudden cardiac death
`Fatal Ml
`Non-fatal Ml
`
`Unstable angina
`CABGorPTCA
`Combined endpoint
`
`Coronary death or Ml
`Fatal Ml or nonfatal Ml
`
`Coronary death
`
`Non-fatal coronary events
`
`Other analyses
`
`Stroke
`
`lschaemic
`
`Haemorrhagic
`
`Other type or not determined
`All-cause death
`
`Number{%) of patients
`EPA
`Control
`
`pvalue Hazard ratio (95 %Cl)
`
`324 (3 5)
`
`262 (2 8)
`
`0·011
`
`0 81 (0 69-0 95)
`
`---•-----
`
`-
`
`·---•--------
`-
`
`------------------------------------
`
`--------------------
`
`-
`
`·------111--------
`
`-------------------- ·---------------------------------------------
`
`--------------------------
`
`----------.--------r-------
`
`17 (0 2)
`14 (02)
`83 (0 9)
`193 (21)
`222 (24)
`
`113 (12)
`93 (10)
`31(03)
`297 (3 2)
`
`18 (0 2)
`11(01)
`62 (0 7)
`147 (16)
`191 (n)
`
`88(09)
`71 (08)
`29 (0 3)
`240 (2 6)
`
`0854
`0557
`0·086
`0·014
`0·135
`
`0·083
`0·091
`0·812
`0·015
`
`106 (0 55-207)
`079 (0 36-174)
`075(054-104)
`076 (0 62-0 95)
`0·86 (071-1·05)
`
`078 (0 59-103)
`077 (0·56-1·05)
`0 94 (0 57-156)
`0 81(068-0 96)
`
`127 (17)
`
`104 (14)
`
`0·132
`
`082 (063-106)
`
`4(01)
`6(01)
`45 (06)
`70 (09)
`74 (10)
`
`55 (07)
`51 (0 7)
`10(01)
`119 (16)
`
`5(01)
`6(01)
`36 (0 5)
`59 (08)
`64 (0 9)
`
`45 (06)
`40 (05)
`11(01)
`95 (13)
`
`0736
`0995
`0.321
`0 338
`0400
`
`Oj22
`0·253
`0·822
`0·102
`
`125 (0 34-4 67)
`1·00 (Oj2-3·11)
`a.so (0.52-124)
`085(060-119)
`0 87 (0 62-121)
`
`0 82 (055-122)
`079 (0 52-119)
`110 (047-2 60)
`080(061-105)
`
`197(107)
`
`158 (87)
`
`0·048
`
`0 81(066-100)
`
`13 (0 7)
`8(04)
`38 (21)
`123 (67)
`148 (8 0)
`
`58 (3 2)
`42 (2 3)
`21 (11)
`178 (97)
`
`13 (0 7)
`5 (0 3)
`26 (14)
`88(48)
`127 (7 0)
`
`43 (24)
`31 (17)
`18 (1 0)
`145 (8 0)
`
`162 (17)
`123 (13)
`39(04)
`5(01)
`265 (2 8)
`
`166 (18)
`115 (12)
`49 (05)
`2 (<0·1)
`286 (31)
`
`0967
`0421
`0·150
`0·019
`0·243
`
`0·156
`0·223
`0 667
`0·080
`
`0785
`0·632
`0.272
`0·252
`
`0 333
`
`1·02 (047-i.19)
`0 64 (0 21-194)
`070 (042-114)
`072 (0 55-0 95)
`0 87 (0 69-110)
`
`075 (051-112)
`075 (047-119)
`0 87 (0 46-1 64)
`082 (066-102)
`
`1·02 (0·91-113)
`0 97 (0 85-110)
`112 (0·91-1.39)
`063 (024-137)
`109 (0 92-128)
`
`0
`
`Figure3: Estimated hazard ratios of clinical endpoints stratified by prevention stratum
`Ml=myocardial infarction. CABG=coronary-artery bypass grafting. PTCA=percutaneous transluminal coronary angioplasty. CAD=coronary-artery disease.
`
`Favours EPA
`
`Favours control
`
`discontinuation because of treatment-related adverse
`events were 1087 (11·7%)
`in the EPA group and
`673 (7 -2%) in the con1rol group. Most adverse effects
`attributable to EPA allocation were regarded as mild.
`The following factors were more common in the EPA
`group than in controls: abnormal laboratory data;
`
`gastrointestinal disturbances such as nausea, diarrhoea,
`or epigastric discomfort; skin abnormalities such as
`eruption,
`itching,
`exanthema, or eczema;
`and
`haemorrhages such as cerebral and fundal bleedings,
`epistaxis, and subcutaneous bleeding. The frequency of
`new cancers did not differ.
`
`1094
`
`www.thelancet.com Vol 369 March 31, 2007
`
`ICOSAPENT DFNDTS00007161
`
`Hikma Pharmaceuticals
`
`IPR2022-00215
`
`Ex. 1005, p. 5 of 9
`
`

`

`Articles I
`
`Control {n=9319}
`
`EPA (n=9326)
`
`Hazard ratio (95% Cl)
`
`Interaction p
`
`117/4380 (2·7J
`
`87/4275 (2·0J
`
`'07/4939 (4·2)
`
`175/5051 (3'5J
`
`016 (0·57-1·00J
`0·84 (0 68-1·02)
`
`0·57
`
`Age(years)
`
`<61
`,61
`
`Sex
`
`Female
`
`126/6411 (2·0J
`
`109/6375 (HJ
`
`0·87 (0·68-H3J
`
`0·43
`
`Male
`
`BMI
`
`<24
`
`<::24
`
`Previous CAD
`
`Absent
`
`Present
`
`Smoking
`
`198/2908 (6·8J
`
`153/2951 (5·2J
`
`Q.76 (0·62-0·94J
`
`136/4404 (HJ
`148/4021 (37)
`
`109/4386 (2·5J
`123/4078 (3·0)
`
`0·80 (0·62-1·03J
`
`0 82 (0·65-105)
`
`0·88
`
`127/7478 (1·7J
`
`104/7503 (1·4 J
`
`0.82 (0·63-1·06J
`
`0.95
`
`197/1841 (10·7J
`
`158/1823 (8·7J
`
`0·81 (0·66-1·00J
`
`Non-smoker
`
`216/7619 (2·8J
`
`170/7496 (2·3J
`
`0·80 (0·66-0·98)
`
`0·89
`
`Smoker
`
`Diabetes
`
`Absent
`
`Present
`
`Hypertension
`
`Absent
`
`108/1700 (6·4J
`
`92/1830 (5·0J
`
`o.y8 (0·59-1·04J
`
`221/7795 (2·8J
`
`175/7810 (2·2J
`
`0·79 (0·65-0·96)
`
`0·62
`
`103/1524 (6·8J
`
`87/1516 (5·7J
`
`0·86 (0·65-H5J
`
`0·85 (0·68-1'06)
`
`Discussion
`Our results show that EPA treatment reduced the
`frequency of major coronary events. The composite
`frequency of the primary endpoint in all patients for the
`EPA group was 19% lower than in controls. The risks of
`unstable angina and non-fatal coronary events were
`also substantially reduced, by 24% and 19%, respectively.
`The beneficial effects of EPA seemed much the same in
`both
`the secondary prevention and
`the primary
`prevention subgroups, although they were significant
`only in the EPA group because of greater numbers of
`events.
`We showed that the reduced risk associated with EPA
`treatment was confined to non-fatal coronary events.
`However, the reduced risk did not apply to coronary
`death or sudden cardiac death in any of our study
`populations or secondary prevention subgroup studies.
`This finding differs from the results of previous
`interventional and observational
`studies. 29 Most
`observational studies report that fish intake only once
`or twice a week or a small intake of fish about 30-60 g
`per day is associated with a 30-60% reduction in the
`risk of fatal coronary events or sudden cardiac deaths,
`but not of non-fatal coronary events.U-5
`7 Secondary
`·
`prevention trials for coronary heart disease report that a
`modest intake of fatty fish
`(200-400 g/week) or
`supplemental intake of EPA plus DHA (1 g/d) reduces
`coronary mortality by about 20-30% in patients who
`9 Experimental
`have already had a myocardial infarction. 8
`·
`and epidemiological studies suggest that fish oil at low
`doses might prevent sudden cardiac death by an
`antiarrhythmic effect.30
`Our findings accord with a cohort study by the Japan
`Public Health Centre, which used a food-frequency
`questionnaire. 31
`Iso and co-workers 31 reported that,
`compared with a small intake of fish (once a week or
`about 20 g per day), a high intake (eight times per week,
`or about 180 g per day) was associated with a substantially
`reduced risk of coronary heart disease, especially non(cid:173)
`fatal cardiac events, in middle-aged Japanese men and
`women. This finding suggests that two protective
`mechanisms of EPA or n-3 polyunsaturated fatty acids
`affect the risk of coronary events: reduction of mortality
`from coronary artery disease and sudden cardiac death
`with a low intake of n-3 polyunsaturated fatty acid, and
`reduction of all coronary events with a high intake of
`n-3 polyunsaturated fatty acids. Our patients could
`possibly all have had intakes of fish that were above the
`threshold for prevention of fatal coronary events or
`sudden cardiac death.4 One potential explanation for
`the strong inverse association with non-fatal coronary
`events in our study population, but not in other study
`populations of non-Japanese patients, is that EPA might
`affect risk only at very high levels of fish intake, such as
`those common in Japan.
`n-3 polyunsaturated fatty acids have antiarrhythmic
`effects and other beneficial effects,3233 such as reduced
`
`167/6037 (2·8J
`
`139/5997 (2·3J
`
`0·57
`
`Present
`
`157/3282 (4·8J
`
`123/3329 (HJ
`
`0·77 (0.61-0·97J
`
`Total cholesterol (mmol/L)
`
`<7·0
`
`~70
`
`167/4700 (3·6J
`
`145/4751 (3·1J
`
`0·86 (0·69-1·08)
`
`0'46
`
`156/ 4608 (3AJ
`
`117/4550 (2·6J
`
`Q.76 (0·60-0·97J
`
`Triglyceride (mmol/L)
`
`<1·7
`
`;::.1.Y
`
`130/4555 (2·9J
`
`188/46,\8 (4·0)
`
`105/4635 (2·3J
`153/!.1563 (3·{l
`
`0·79 (0.61-i.G2J
`
`0 84 (0·68-1·04)
`
`HDL-cholesterol (mmol/L)
`
`<1·5
`
`~l·S
`
`206/4316 (4·8J
`
`91/4285 (2·1)
`
`154/4149 (HJ
`91/4491 (2·0J
`
`0·78 (0·64-0·96)
`0·96 (0 72-i.28)
`
`LDL-cholesterol (mmol/L)
`
`0·75
`
`0·26
`
`<4·7
`
`~4.Y
`
`129/4160 (3·1J
`
`108/4251 (2·5J
`
`0.82 (0·64-1·06)
`
`0.83
`
`156/ 4157 (3'8)
`
`131/4097 (3·2)
`
`0·86 (0·68-1·08)
`
`Data are number of patients(%) or hazard ratio (95% Cl). p values are fort he test of heterogeneity. CAD,,,coronary
`artery disease. BMl=body-mass index, which is weight in kg divided by the square of height in metres. LDL==low-density
`lipoprotei n. H DL=high-density lipoprotei n.There is a deficit of clinical data in some patients with the events.
`
`Table 2: Subgroup analysis
`
`10
`
`0
`
`-10
`
`,o~
`~
`~
`c' "'' ~ -20
`u
`
`-30
`
`-40
`
`Patients with
`~1·7mmo!/L
`
`TC
`
`LDL-C
`
`HDL-C
`
`Triglyceride
`
`Triglyceride
`
`Figure4: Percentage changes from baseline in serum lipid profile
`TC=total cholesterol. LDL (=low-density lipoprotein cholesterol. HDL (=high-density lipoprotein cholesterol.
`
`www.thelancet.com Vol 369 March 31, 2007
`
`1095
`
`ICOSAPENT DFNDTS00007162
`
`Hikma Pharmaceuticals
`
`IPR2022-00215
`
`Ex. 1005, p. 6 of 9
`
`

`

`I Articles
`
`Total number of adverse experiences(%)
`
`2004(21·7%)
`
`2334 (25·3%)
`
`<0·0001
`
`Control (n=9319)
`
`EPA (n=9326)
`
`pvalue
`
`Newly diagnosed cancer
`
`Total
`
`Stomach
`
`Lung
`
`Colorectal
`
`Breast
`
`218(2-4%)
`
`242 (2·6%)
`
`37(0·4%)
`
`37(0·4%)
`
`29 (0·3%)
`
`21 (0·2%)
`
`53 (0·6%)
`
`32 (0·3%)
`
`26 (0·3%)
`
`16 (0·2%)
`
`0·26
`
`0·09
`
`0·54
`
`0·68
`
`0-41
`
`Common adverse experiences
`
`Pain (joint pain, lumbar pain, muscle pain)
`
`180 (2·0%)
`
`Gastrointestinal disturbance (nausea,
`diarrhoea, epigastric discomfort)
`
`Skin abnormality (eruption, itching,
`exanthema, eczema)
`
`Haemorrhage (cerebral, fundal, epistaxis,
`subcutaneous)
`
`Abnormal laboratory data
`
`Total
`
`CPK increased
`
`GOT increased
`
`Sugar blood level increased
`
`155(1'7%)
`
`144 (1·6%)
`
`352 (3·8%)
`
`0.04
`
`<0·0001
`
`65(0·7%)
`
`160(H%)
`
`<0·0001
`
`60(0·6%)
`
`105 (l-1%)
`
`0·0006
`
`322 (3·5%)
`
`116 (1·2%)
`
`38 (0·4%)
`
`27 (0·3%)
`
`378 (4·1%)
`
`126(1-4%)
`
`59 (0·6%)
`
`38 (0·4%)
`
`0.03
`
`0·52
`
`0·03
`
`0·17
`
`fish is about five times higher than that in other
`countries. 28 We did not use a food-frequency question(cid:173)
`naire to measure fish intake; instead, at baseline, we
`measured plasma fatty acid concentrations that indicate
`fish consumption and EPA intake. Plasma EPA was
`2 · 9 mol% at baseline in our study population, which is
`similar to reports by !so and co-workers40 that serum
`EPA composition was 4· 1 mol% in rural Japanese and
`2 · 4 mol% in urban Japanese; these values are much
`higher than those recorded in the USA,