Design and Baseline Characteristics of the Incremental Decrease in End Points Through Aggressive Lipid Lowering Study Terje R. Pedersen, MD, Ole Faergeman, MD, John J.P. Kastelein, MD, Anders G. Olsson, MD, Matti J. Tikkanen, MD, Ingar Holme, PhD, Mogens Lytken Larsen, MD, Fredrik S. Bendiksen, MD, Christina Lindahl, MD, and Gary Palmer, MD, on behalf of the Incremental Decrease in End Points Through Aggressive Lipid Lowering Study Group The Incremental Decrease in End Points through Aggressive Lipid Lowering (IDEAL) study is an investigatorinitiated trial designed to determine whether additional clinical benefit might be gained through a strategy that decreases levels of low-density lipoprotein cholesterol levels better than those currently achieved with established statin therapy in patients who have coronary heart disease. IDEAL is a multicenter prospective, randomized, open-label, blinded, end point classification study. Patients who had myocardial infarction were randomized to prescription treatment with 80 mg/day of atorvastatin or 20 mg/day of simvastatin (the dose was increased to 40 mg/day at week 24 in those patients whose plasma total cholesterol remained >5.0 mmol/L, or 190 mg/dl, or whose low-density lipoprotein cholesterol remained >3.0 mmol/L, or 115 mg/dl). The primary clinical outcome variable is the time to initial occurrence of a major coronary event, which is defined as

nonfatal acute myocardial infarction, coronary death, or resuscitated cardiac arrest. The study is designed to have a power of 90% to detect a relative decrease of 20% in the atorvastatin-group compared with the simvastatin-group in the number of major events caused by coronary heart disease over ⬃5.5 years. The 8,888 randomized patients had the following characteristics: mean age 61.7 ⴞ 9.5 years, 19.1% women (mean age 64.0 ⴞ 9.5 years), baseline total cholesterol 5.1 ⴞ 1.0 mmol/L (197 mg/dl), low-density lipoprotein cholesterol 3.2 ⴞ 0.9 mmol/L (124 mg/dl), and high-density lipoprotein cholesterol 1.2 ⴞ 0.3 mmol/L (46 mg/dl). Drug treatment before randomization consisted of statins in 77% of patients, aspirin in 78.9%, ␤ blockers in 75.1%, and angiotensin-converting enzyme inhibitors in 30%. 䊚2004 by Excerpta Medica, Inc. (Am J Cardiol 2004;94:720 –724)

reater percent decreases in levels of low-density lipoprotein cholesterol (LDL) than those G achieved in the Scandinavian Simvastatin Survival

higher LDL cholesterol concentrations.2 These data were attributed to a mean difference in LDL cholesterol levels of ⬃0.9 to 1.0 mmol/L (35 to 39 mg/dl) between the treated and control groups. In contrast, the 4S achieved a mean difference of 1.7 mmol/L (66 mg/dl) between the statin-treated and control groups. The Incremental Decrease in End Points through Aggressive Lipid Lowering (IDEAL) study intends to achieve an additional interventional decrease in LDL cholesterol of 0.6 to 0.7 mmol/L (23 to 27 mg/dl) by therapy with 80 mg of atorvastatin and therefore compares the effect of a mean decrease of ⬃55% in LDL cholesterol concentrations with a 35% decrease compared with levels before statin therapy. The beneficial effects of a 35% decrease in LDL cholesterol with 10 mg/day of atorvastatin on clinical end points was recently proved in the Anglo-Scandinavian Cardiac Outcomes Trial Lipid-Lowering Arm.3 The more recent Pravastatin or Atorvastatin Evaluation and Infection Therapy-Thrombolysis in Myocardial Infarction-22 (PROVE IT-TIMI 22) trial showed that 80 mg of atorvastatin provides greater protection than the less potent 40 mg of pravastatin in patients who have had a recent acute coronary syndrome.4

Study (4S)1 might yield greater benefit in patients who have coronary heart disease. The Heart Protection Study has provided evidence that treatment with 40 mg of simvastatin produces a similar relative decreased risk in patients with LDL cholesterol levels below currently recommended target levels compared with the risk observed among patients presenting with

From the Center for Preventive Medicine, Ullevål University Hospital, Oslo, Norway; the Department of Medicine-Cardiology A, Århus University Hospital, Århus, Denmark; the Academic Hospital Amsterdam, Amsterdam, The Netherlands; the Department of Internal Medicine, University Hospital, Linköping, Sweden; the Medical Clinic, Helsinki University Hospital, Helsinki, Finland; the Hamar, Lysaker, Norway; Pfizer Sweden, Täby, Sweden; and Pfizer Inc., New York, New York. This study was supported by Pfizer Inc., New York, New York. Manuscript received January 15, 2004; revised manuscript received and accepted June 7, 2004. Address for reprints: Terje R. Pedersen, MD, Center for Preventive Medicine, Building K, Ullevål University Hospital, N-0407 Oslo, Norway. E-mail: [email protected].

720

©2004 by Excerpta Medica, Inc. All rights reserved. The American Journal of Cardiology Vol. 94 September 15, 2004

0002-9149/04/$–see front matter doi:10.1016/j.amjcard.2004.06.003

METHODS

Study aim: The primary objective of the IDEAL

study is to investigate whether an incremental decrease in the risk of coronary heart disease can be achieved by a greater decrease in LDL cholesterol in patients who use secondary prevention than that attained with a treatment strategy representing the current best clinical practice as established in the 4S. The time to occurrence of a major coronary event will be compared in 2 groups of patients: 1 that achieves an average decrease of ⬃55% and 1 that achieves an average decrease of ⬃35% in LDL cholesterol. Study eligibility: Men and women ⱕ80 years old who had been hospitalized or had a history of a definite myocardial infarction and who qualified for statin therapy according to guidelines at the time of recruitment were eligible for study entry. Major exclusion criteria were a known hypersensitivity to any statin; active liver disease or hepatic dysfunction defined as a level of alanine aminotransferase or aspartate aminotransferase ⬎2 times the upper limit of normal; women who were pregnant or breast-feeding; nephrotic syndrome; uncontrolled diabetes mellitus; uncontrolled hypothyroidism; plasma triglycerides ⬎6.8 mmol/L (600 mg/dl); congestive heart failure (New York Heart Association classification IIIb or IV); hemodynamically important valvular disease; gastrointestinal disease that might limit drug absorption or partial ileal bypass; other significant abnormalities that in the investigator’s opinion could compromise the patient’s safety or successful participation in the study; and treatment with medications, such as other lipid-lowering drugs or any immunosuppressive agent that could compromise the patient’s safety or obscure the efficacy or safety comparisons. Patients already titrated to a dose of statin more than the equivalent of 20 mg/day of simvastatin were also excluded. Study design, treatment, and patient management:

IDEAL is a multicenter, randomized, open-label study with stratified randomization carried out in blocks by center and with blinded end point classification (the so-called PROBE design5). After dietary counseling, all patients fulfilling the eligibility criteria were randomized to treatment with 80 mg/day of atorvastatin or 20 mg/day of simvastatin. No drug-free period was required. Study medication was provided by prescription in all countries except Finland, where it was dispensed at the clinical centers at the expense of the sponsor. In Norway and Sweden, patients paid part of the cost of medication; in Denmark and The Netherlands, all costs were reimbursed. These procedures corresponded to reimbursement rules for statin therapy in the respective countries. The difference in cost between simvastatin and atorvastatin was equalized by the sponsor through agreements with governmental agencies in these countries. The study protocol allowed the dose of simvastatin to be increased to 40 mg/day at week 24 in those patients whose plasma total cholesterol remained ⬎5.0 mmol/L (190 mg/dl). All patients were followed up 12 and 24 weeks after randomization and will be

followed every 6 months thereafter until the predetermined number of primary end points for the study is observed. This is estimated to occur after a median duration of 5.5 years. Follow-up visits occur in the outpatient clinic of the recruiting hospital or a specialist center to monitor compliance and adverse events and study end points. Baseline serum lipid and lipoprotein levels were measured from fasting blood samples taken 7 to 14 days before randomization. Lipid profiles were determined at weeks 12 and 24 and 1 year and will be measured every 12 months thereafter and at the final visit. Liver enzymes (specifically aspartate aminotransferase and alanine aminotransferase), creatine phosphokinase, and other laboratory measurements were determined at randomization and at weeks 12 and 24 and will be determined every 12 months thereafter and at the final visit. All laboratory procedures and lipoprotein determinations are conducted at a central laboratory. Results of lipid and lipoprotein determinations will not be revealed to investigators or other study personnel during the study with exception of critical values of total and/or LDL cholesterol values at week 24, when titration of the simvastatin dose was allowed (as previously mentioned). Patient hospital records are marked with labels telling the staff not to measure such values. General physical examinations and electrocardiograms are carried out annually. Study management: The study was initiated by the investigators and is scientifically led by a steering committee consisting of independent researchers and investigators and 3 members employed by the sponsor. This committee has the full responsibility for the study protocol and publications from the study. The committee is granted full access to all data at study completion through a contract with the sponsor. The study is monitored by employees of the sponsor who collect all case report forms after review and submit them to a contract research organization (COVANCE, Princeton, New Jersey) that enters the data into a database. Data are then reviewed for errors and inconsistencies through a comprehensive system of algorithms, and queries are sent to investigators for clarifications. The contract research organization is responsible for preparing interim analysis reports for the data safety and monitoring board. These reports are not shared with the sponsor or any other personnel in the trial. Study outcomes: The primary clinical outcome variable is the time to first occurrence of a major coronary event, which is defined as nonfatal acute myocardial infarction according to criteria of the Joint European Society of Cardiology/American College of Cardiology, coronary death, or resuscitated cardiac arrest. Secondary outcomes are the occurrence of any primary event, any coronary artery revascularization procedure, or hospitalization for unstable angina; hospitalization with a primary diagnosis of congestive heart failure; any cerebrovascular event; peripheral arterial disease; any cardiovascular event; and all-cause mortality rate. Because all end points except death and new diagnosis of peripheral artery disease require

CORONARY ARTERY DISEASE/DESIGN AND BASELINE DATA IDEAL STUDY

721

hospitalization, the potential for biased reporting by investigators is being minimized by study monitors who review patient records at regular intervals to search for potential end points. An end point committee blindly and individually reviews all primary and secondary end points to confirm that the data support the end point designation. Any differences between committee members are adjudicated at regular meetings. To ensure blinding by this committee, all end point reports accompanied by narratives or copies of hospital records are first sent to an independent center for masking. Data safety monitoring board: An independent safety data monitoring board, not otherwise involved in the conduct of the trial, will monitor aggregated end point summaries and medically serious adverse events on an ongoing basis. The primary responsibility of this board is the safety of trial participants. The board will convene every 6 months to review the status of the study and has the power to make recommendations to halt the trial, continue the trial beyond the planned size or duration, or change trial procedures. Sample size calculation: The trial originally planned recruitment of 7,600 patients based on experience from the 4S, in which the mean LDL cholesterol at baseline was ⬃4.87 ⫾ 0.65 mmol/L. In contrast to the 4S, in which patients were randomized on average ⬎3 years after their first myocardial infarction or first diagnosis of angina pectoris, the plan was to randomize patients participating in the IDEAL study within 6 weeks of a myocardial infarction. Because of competition with other ongoing trials in this population, most patients had been recruited several months or years after the initial myocardial infarction. As a consequence, most patients were on lipid-lowering therapy, and the profile of the first patients recruited made it more likely that the sample size calculations should be based on those of the Cholesterol and Recurrent Events (CARE) study.6 In that study, the mean baseline LDL cholesterol level was 3.6 mmol/L (139 mg/ dl). It is expected that 20 to 40 mg/day of simvastatin will decrease LDL cholesterol levels by 35% and that 80 mg/day of atorvastatin will decrease LDL cholesterol levels by 55% compared with nontreatment levels. This difference in turn is expected to decrease the number of 5-year recurrent coronary events by 20% in the atorvastatin group compared with the simvastatin group. These figures reflect the decreases achieved in long-term statin trials after taking into account medication switch-over, decreased compliance, and number of dropouts. Therefore, assuming a 10% recurrent event rate for 20 to 40 mg/day of simvastatin (based on experience from the 40-mg pravastatin group in the CARE trial) and a 7.9% recurrent event rate for 80 mg/day of atorvastatin, it was determined that 7,764 patients would provide 90% power to detect a 21% decrease in recurrent coronary events over 5 years between treatment groups when using a 2-tailed test (␣ ⫽ 0.05) based on the log-rank test. After adjusting for interim observations at 50% and 75% information times (percent primary end points) using O’BrienFleming boundaries, the trial randomized 8,888 pa722 THE AMERICAN JOURNAL OF CARDIOLOGY姞

VOL. 94

tients with a target of 774 patients with primary events. Interim analyses: Interim analyses for IDEAL will be performed when ⬃50% and ⬃75% of primary events have occurred, and they will include assessment of recruitment, subject characteristics, primary efficacy and safety variables, all-cause mortality rates, and adverse events. Early discontinuation of the trial is a complex decision involving a number of factors aside from the simple comparison of 1 statistic with 1 critical value. A recommendation to modify or terminate the trial would not be based solely on the crossing of a group-sequential boundary. Baseline data: RECRUITMENT: Enrollment into the IDEAL study was carried out at 190 centers in Norway, Sweden, Finland, Denmark, Iceland, and The Netherlands. The first patient was recruited on March 31, 1999 and the final patient was randomized on March 29, 2001. Of the 9,689 patients screened, 8,888 were randomized to open-label prescription treatment with 80 mg/day of atorvastatin or 20 mg/day of simvastatin. At the 6-month follow-up, the simvastatin dose was increased to 40 mg in 20.3% of patients. BASELINE CHARACTERISTICS: Selected baseline characteristics of randomized patients are listed in Table 1. Most patients were randomized ⬎6 months after their myocardial infarction. Accordingly, 77% were receiving statin therapy at the time of randomization. Serum concentrations at baseline of different lipid parameters, glucose, and glycosylated hemoglobin are presented in Table 2. Because patients already titrated to larger doses of statin were excluded and most randomized patients were on statin therapy, the mean baseline LDL cholesterol concentration was only 3.2 mmol/L (122 mg/dl).

DISCUSSION Post hoc analysis of the 4S data demonstrated a direct relation between the degree to which LDL cholesterol was decreased and the risk of future clinical events.7 Therefore, in the IDEAL study, we are using 80 mg of atorvastatin to determine whether more aggressive lipid lowering would improve the prognosis in patients who have coronary heart disease compared with a strategy that reflects the current best clinical practice resulting from the 4S. Treatment in the present study is administered in an open-label manner. The validity of the PROBE design is well documented in published reports, and a significant bias that affects the end points of the trial is highly unlikely.5 Patients were assigned to treatment by central randomization, thus effectively eliminating the potential for biased assignment. Clearly, the treatment groups in the IDEAL study differ from each other in 2 respects, drug type and intensity of therapy. Although our primary objective is not to compare 2 lipid-lowering agents, atorvastatin and simvastatin, it is evident that, irrespective of the trial results, we will be faced with ⬎1 explanation for any differences in outcome that might occur. For this reason, IDEAL has been closely linked to the Treating to New Targets (TNT) study.8 In contrast to the SEPTEMBER 15, 2004

IDEAL study, TNT will evaluate the difference in incidence of end points at 2 different absolute levels of treatment of LDL cholesterol among patients who have coronary heart disease. In that study, patients are being treated with 2 different doses of atorvastatin, 10 mg and 80 mg. The first group is expected to reach a mean LDL cholesterol level of ⬃2.6 mmol/L (100 mg/dl), and the second group is expected to reach of level of 2.0 mmol/L (75 mg/dl). The close collaboration between IDEAL and TNT means that together these studies will provide insight into the importance of the percentage by which cholesterol is decreased and the absolute level of LDL cholesterol that can be achieved for the prevention of cardiovascular events.

TABLE 1 Baseline Characteristics of Patients (n ⫽ 8,888)

Men (age) Women (age) Age (yrs) ⬍40 40–44 45–49 50–54 55–59 60–64 65–69 70–74 ⬎75 Caucasian Noncaucasian No. of previous myocardial infarctions Yrs since last myocardial infarction Coronary angioplasty only Coronary bypass only Angioplasty and bypass Current smoker Ex-smoker Systemic hypertension history Diabetes mellitus Body mass index ⬎30 kg/m2 Cerebrovascular disease Peripheral vascular disease Congestive heart failure Concomitatant treatment Aspirin Other antiplatelet drugs Anticoagulants ␤ Blockers Calcium antagonists ACE inhibitors Angiotensin II receptor blockers Digitalis Spironolactone Other diuretics Nitrates Hormone replacement therapy Simvastatin Atorvastatin Pravastatin Fluvastatin Lovastatin Cerivastatin Nonstatin drugs Systolic blood pressure (mm Hg) Diastolic blood pressure (mm Hg) Weight (kg)

Mean ⫾ SD

n (%)

61.2 ⫾ 9.4 64.0 ⫾ 9.5 61.7 ⫾ 9.5

7,187 (80.9%) 1,701 (19.1%)

1.2 ⫾ 0.5 3.9 ⫾ 4.9

136.9 ⫾ 20.1 80.4 ⫾ 10.2 82.5 ⫾ 13.8

87 263 601 1,193 1,448 1,537 1,654 1,318 785 8,816 72

(1.0%) (3.0%) (6.8%) (13.4%) (16.3%) (17.3%) (18.6%) (14.8%) (8.8%) (99.2%) (0.8%)

1,764 1,477 289 1,832 5,197 2,857 1,057 1,826 650 355 1,169

(19.9%) (16.6%) (3.3%) (20.6%) (58.5%) (32.1%) (11.9%) (20.7%) (7.3%) (4.0%) (13.2%)

7,017 151 1,022 6,677 1,723 2,668 526 280 182 1,662 3,361 185 4,537 1,025 872 253 107 52 48

(78.9%) (1.7%) (11.5%) (75.1%) (19.4%) (30.0%) (5.9%) (3.2%) (2.0%) (18.7%) (37.8%) (10.9% of women) (51.0%) (11.5%) (9.8%) (2.8%) (1.2%) (0.6%) (0.5%)

ACE ⫽ angiotensin-converting enzyme.

TABLE 2 Fasting Serum Levels of Lipids, Glucose, and Glycosylated Hemoglobin Concentration (mean ⫾ SD) mg/dl Total cholesterol HDL cholesterol LDL cholesterol Triglycerides Total/HDL ratio Apolipoprotein A-1 (g/L) Apolipoprotein B (g/L) Hemoglobin A1c (%) Glucose HDL ⫽ high-density lipoprotein.

196 46 122 149 4.5 1.4 1.2 5.0 107.3

⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾

39.3 12.0 34.7 78.5 1.4 0.2 0.3 0.9 29.9

mmol/L ⫾ 1.0 ⫾ 0.3 ⫾ 0.9 ⫾ 0.9 — — — — 6.0 ⫾ 1.7

5.1 1.2 3.2 1.7

Acknowledgment: We are indebted to the departments of health and the national health insurance institutions in Norway and Sweden, the National Medicines Agency, Copenhagen, Denmark, and the national societies of pharmacies in these countries and in Iceland for their cooperation in organizing the reimbursement and study drug delivery systems for the trial.

APPENDIX Steering Committee: Terje R. Pedersen, MD (chair); Ole Faegeman, MD; Ingar Holme, PhD; Anders G. Olsson, MD; Matti Tikkanen, MD; John J.P. Kastelein, MD; Mogens Lytken Larsen, MD; Christina Lindahl, MD; Gary Palmer, MD; Fredrik S. Bendiksen, MD; Sverre Maehlum, MD. Data Safety and Monitoring Board: Desmond G. Julian, MD (chair); Hans Wedel, PhD; Knut Rasmussen, MD; Christie Ballantyne, MD. End-Point Committee: Kristian Thygensen, MD (chair); Bengt W. Johansson, MD; Per Lund-Johansen, MD. 1. Scandinavian Simvastatin Survival Study Group. Randomised trial of cholesterol lowering in 4444 patients with coronary heart disease: the Scandinavian Simvastatin Survival Study (4S). Lancet 1994;344:1383–1389. 2. Heart Protection Study Collaborative Group. MRC/ BHF Heart Protection Study of cholesterol lowering with simvastatin in 20 536 high-risk individuals: a randomised placebo-controlled trial. Lancet 2002;360:7–22. 3. Sever PS, Dahlöf B, Poulter NR, Wedel H, Beevers G, Caulfield M, Collins R, Kjeldsen SE, Kristinsson A, Mcinnes GT, et al. Prevention of coronary and stroke events with atorvastatin in hypertensive patients who have average or lower-than-average cholesterol concentrations, in the Anglo-Scandinavian Cardiac Outcomes Trial— Lipid Lowering Arm (ASCOT-LLA): a multicentre randomized controlled trial. Lancet 2003;361:1149 –1158. 4. Cannon CP, Braunwald E, McCabe CH, Rader DJ, Rouleau JL, Belder R, Joyal SV, Hill KA, Pfeffer MA, Skene AM. Comparison of intensive and moderate lipid lowering with statins after acute coronary syndromes. N Engl J Med 2004;350:1495–1504. 5. Hansson L, Hedner T, Dahlöf B. Prospective Randomized Open Blinded End-point (PROBE) study. A novel design for intervention trials. Blood Press 1992; 1:113–119.

CORONARY ARTERY DISEASE/DESIGN AND BASELINE DATA IDEAL STUDY

723

6. Sacks FM, Pfeffer MA, Moye LA, Rouleau JL, Rutherford JD, Cole TG, Brown L, Warnica JW, Arnold JMO, Wun C, et al. The effect of pravastatin on coronary events after myocardial infarction in patients with average cholesterol levels. N Engl J Med 1996;335:1001–1009. 7. Pedersen TR, Olsson AG, Faergeman O, Kjekshus J, Wedel H, Berg K, Wilhelmsen L, Haghfelt T, Thorgeirsson G, Pyörälä K, et al. Lipoprotein changes

724 THE AMERICAN JOURNAL OF CARDIOLOGY姞

VOL. 94

and reduction in the incidence of major coronary heart disease events in the Scandinavian Simvastatin Survival Study (4S). Circulation 1998;97:1453–1460. 8. Waters DD, Guyton JR, Herrington DM, McGowan MP, Wenger NK, Shear C. Treating to New Targets (TNT) study: does lowering low-density lipoprotein cholesterol levels below currently recommended guidelines yield incremental clinical benefit? Am J Cardiol 2004;93:154 –158.

SEPTEMBER 15, 2004