Author + information
- Received July 15, 2015
- Accepted August 18, 2015
- Published online October 20, 2015.
- Michael J. Domanski, MD∗,†,‡∗ (, )
- Valentin Fuster, MD, PhD∗,§,
- Francisco Diaz-Mitoma, MD, PhD‖,
- Scott Grundy, MD, PhD¶,
- Donald Lloyd-Jones, MD#,
- Muhammad Mamdani, PharmD, MPH∗∗,
- Robin Roberts, MSc††,
- Kevin Thorpe, MSc∗∗,
- Judith Hall, PhD∗∗,
- Jacob A. Udell, MD†,‡,‡‡ and
- Michael E. Farkouh, MD, MSc∗,†,‡
- ∗Icahn School of Medicine at Mount Sinai, New York, New York
- †Heart and Stroke/Richard Lewar Centre of Excellence, University of Toronto, Toronto, Ontario, Canada
- ‡Peter Munk Cardiac Centre, University of Toronto, Toronto, Ontario, Canada
- §Centro Nacional de Investigaciones Cardiovasculares Carlos III (CNIC), Madrid, Spain
- ‖Advanced Medical Research Institute of Canada, Sudbury, Ontario, Canada
- ¶University of Texas, Southwestern University Medical School at Dallas, Dallas, Texas
- #Northwestern University Medical School, Chicago, Illinois
- ∗∗Applied Health Research Centre, Li Ka Shing Knowledge Institute, University of Toronto, Ontario, Canada
- ††McMaster University, Hamilton, Ontario, Canada
- ‡‡Women's College, Peter Munk Centre, Research Institute and Cardiovascular Division, Department of Medicine, Women's College Hospital, University of Toronto, Toronto, Ontario, Canada
- ↵∗Reprint requests and correspondence:
Dr. Michael J. Domanski, Peter Munk Cardiac Centre, 585 University Avenue, 4N484, Toronto, Ontario M5G 2N2, Canada.
Atherosclerotic cardiovascular disease (ASCVD) events, including coronary heart disease and stroke, are the most frequent cause of death and major disability in the world. Current American College of Cardiology/American Heart Association primary prevention guidelines are mainly on the basis of randomized controlled trials of statin-based low-density lipoprotein cholesterol (LDL-C)–lowering therapy for primary prevention of ASCVD events. Despite the clear demonstration of statin-based LDL-C lowering, substantial 10-year and lifetime risks of incident ASCVD continue. Although the 10-year risk is low in young and middle-aged adults who would not be treated according to current guidelines, they ultimately account for most incident ASCVD. If statin-based LDL-C lowering were initiated in them at an age before complex coronary plaques are common in the population, a substantial reduction in lifetime risk of incident coronary heart disease might be achieved. We examine this hypothesis and introduce the design of a currently recruiting trial to address it. (Eliminate Coronary Artery Disease [ECAD]; NCT02245087)
- hydroxymethylglutaryl-CoA reductase inhibitors
- primary prevention
Atherosclerotic cardiovascular disease (ASCVD), including coronary heart disease (CHD) and ischemic stroke, is the most frequent cause of death in the world (1–4). Even when not lethal, ASCVD frequently results in disability, and the economic price of caring for the patients represents a huge opportunity cost (resources that could have been otherwise expended) for society. Demonstration of a means to achieve a substantial reduction in incident ASCVD would represent a paradigm-shifting advance in public health and would profoundly alter medical practice. Although there have been major advances in revascularization and also in medical therapy for patients with symptomatic ASCVD, many initial events are fatal, treatment of manifest ASCVD palliates but does not cure, and a truly decisive effect on the ASCVD epidemic will require more effective prevention of incident disease.
In this paper, we consider the potential next steps in primary prevention of ASCVD that could improve results in comparison with those expected from utilizing current guidelines. We have previously discussed the possible therapeutic potential of low-density lipoprotein cholesterol (LDL-C) lowering early in life (1). We extend that discussion by addressing the hypothesis that most of the therapeutic potential of lifelong pharmacological LDL-C lowering might be achieved by pharmacological LDL-C lowering in adults who are not yet candidates for guideline-based therapy due to their low calculated risk of incident ASCVD while they are still young enough that complex coronary plaques are infrequent in the population. We also introduce a recruiting, randomized, controlled, primary prevention trial of statin-based LDL-C lowering, now underway, which is designed to address this hypothesis.
Atherosclerotic Plaque, LDL-C, and Guideline-Based ASCVD Event Prevention
Atherosclerotic plaque, either with a superimposed thrombus or by restricting blood supply, is the cause of the vast majority of ASCVD clinical events. If plaque formation or progression could be prevented, the vast majority of incident ASCVD events should also be prevented.
Production of atherosclerotic plaque is a complex process in which LDL-C enters the arterial wall to a degree modulated by the serum LDL-C concentration (1,5). In the arterial wall, LDL-C is chemically modified and taken up by macrophages derived from circulating monocytes, thus instigating plaque formation. As would be expected on the basis of this mechanism, serum LDL-C concentration has been demonstrated to have a strong, graded association with both plaque progression and burden in imaging studies and with clinical events in studies of ASCVD epidemiology (6–11).
These understandings formed the basis for the hypothesis that pharmacological lowering of LDL-C would prevent incident ASCVD. A number of landmark primary prevention clinical trials have been conducted, mostly in those older than 50 years of age, which demonstrated a benefit of pharmacological LDL-C lowering of a magnitude directly related to the LDL-C level achieved with treatment (Central Illustration) (12–18).
Nonetheless, considerable incident ASCVD develops in trial patients treated with LDL-C–lowering medication. The persistence of incident CHD in the clinical trial population, despite statin treatment, could result from a substantial plaque burden already present before treatment. This suggests that if LDL-C–lowering treatment were started earlier in life (and thus earlier in the natural history of a person’s atherosclerotic disease than will occur with the current guidelines), greater (perhaps much greater) inroads into preventing incident ASCVD could be achieved than afforded by treatment according to the current guidelines.
LDL-C Lowering: When and How Much?
Central to further advances in primary prevention is an improved understanding of when in life to start pharmacological LDL-C lowering and how low to drive the serum LDL-C concentration.
Although now rare, hunter-gatherer populations, whose LDL-C throughout their lives was in the 50 to 70 mg/dl range (similar to current neonates), rarely experienced CHD (5,19–23). That the difference between hunter-gatherers and “modern” populations was lower low-density lipoprotein dating from birth, rather than a constitutional difference, is suggested by the fact that CHD became common in these populations as they adopted the diet and lifestyle of “modern” populations (19,24). This leads to the hypothesis that if the hunter-gatherer level of LDL-C could be maintained throughout life in current populations, an incidence of CHD similar to that found in hunter-gatherer populations might be achieved.
Data from the ARIC (Atherosclerosis Risk in Communities) study (25) underscore the preventive potential of lifelong lower LDL-C. The proprotein convertase subtilisin/kexin type 9 (PCSK9) gene codes for a serine protease that degrades LDL-C receptors. Patients with a nonsense mutation of PCSK9 have reduced enzyme activity, resulting in a higher LDL-C receptor density, and thus, a lower serum LDL-C concentration throughout their lives. As a result, carriers of these nonsense gene mutations have a lower risk of ASCVD events than patients with the normal gene. The risk reduction is greater than would be expected from the modest difference in serum LDL-C that results from the mutation (1). This can be seen in the relationship of ASCVD events and serum LDL-C in patients with the PCSK9 nonsense mutation. These patients have a greater decrease in risk associated with a given decrease in LDL-C compared with the risk reduction for statin-treated older patients and for control subjects in the statin trials with the same degree of LDL-C reduction.
Further supporting the notion of greater reduction of incident ASCVD events resulting from earlier initiation of LDL-C lowering is a meta-analysis by Law et al. (26) showing that the CHD risk reduction associated with a given magnitude of LDL-C lowering is negatively associated with increasing age after the fifth decade. Data for subjects younger than 50 years of age were not available, although a qualitatively similar relationship seems likely, particularly because substantial subclinical atherosclerosis is present in many people in their middle years (10). Other recently reported data also suggest that the ASCVD event rate increases with increasing atherosclerosis burden (11).
Also supportive of enhanced effectiveness of earlier LDL-C lowering are data showing that the therapeutic benefit of LDL-C-lowering treatment is maintained over time. Ten years after the original 5-year trial was completed, long after many placebo-treated patients would have had statin treatment initiated and some statin-treated patients would have discontinued statin, data from WOSCOPS (West of Scotland Trial Coronary Prevention Study) (14,27) shows that patients randomized to statin treatment still had a lower rate of CHD events than patients randomized to placebo. In fact, the risk reduction was greater than at the end of the trial. This continuing difference suggests that the earlier LDL-C lowering is started, the more effective the intervention.
Additionally, a meta-analysis of trials by the Cholesterol Treatment Trialists also suggested increasing benefit over time, which comports with the previously mentioned and suggests that benefit over an intermediate term also predicts long-term benefit (28,29).
In aggregate, the available studies suggest that maintaining low serum LDL-C from birth, or from as early as practical, may be the most effective approach for primary prevention of ASCVD. As a corollary, restoring neonatal levels of LDL-C early in life is likely to be highly effective. However, the notion of starting pharmacological intervention early enough to prevent the rise in serum LDL-C that occurs through childhood and adolescence offers the unappealing scenario of starting very young people at very low short-term risk on medication and continuing the medication over a lifetime. Once the atherosclerosis process has started, particularly once it has progressed to the point where complex plaques are present, maximal prevention may require a lower serum LDL-C than would be the case had treatment been started earlier and may not be as effective, no matter how much LDL-C is lowered. However, the association of CHD prevention effect size with the age at which cholesterol lowering occurs at younger ages is unknown. Is there a “sweet spot” of age, perhaps significantly into adulthood, before the disease has progressed to the presence of complex plaques, when pharmacological lowering of LDL-C might result in a reduction in CHD incidence similar to that obtained by maintaining the level common in neonates throughout childhood, adolescence, and adult life? If so, it may be possible to achieve the low lifetime risk of incident CHD (about 5%) that occurs when asymptomatic patients have optimal risk factor control at 50 years of age (5), while avoiding subjecting young people to a lifetime of pharmacological therapy.
In what age range should we search for the “sweet spot” postulated previously? A reasonable approach to finding the latest time at which the effect size would remain similar to that for lifelong low LDL-C would be to hypothesize that a major treatment effect of lowering LDL-C would still be present at an age at which complex plaques are still infrequent. Data from the PDAY (Pathological Determinants of Atherosclerosis in Youth) study suggest that fewer than 5% of patients age 30 years to 34 years have raised coronary lesions (30). Thus, it may be possible to delay pharmacological LDL-C lowering into early- or mid-middle age and still achieve most, perhaps essentially all, of the benefit that attends congenitally low serum LDL-C. A clinical trial to determine whether pharmacological LDL-C lowering in the middle years (before complex lesions are widespread) in patients who would not be treated under current guidelines can reduce incident clinical ASCVD in the overall population beyond that afforded by treatment according to current guidelines is a next logical step in improving primary prevention.
General trial considerations
Any such trial would enter patients who, by virtue of age and/or low calculated 10-year risk, would not qualify under current guidelines for pharmacological lipid lowering. To be practical, the trial would need to avoid requiring far more subjects than the number of patients in current large clinical trials. The endpoints should be hard ASCVD clinical endpoints that are easily ascertained, such as death, myocardial infarction (MI), stroke, and first revascularization.
Also, as a matter of practicality, the trial duration should not exceed about 10 years. Even a reduction in 10-year ASCVD incidence would affect guideline content. Additionally, the known continuing separation of treatment groups over time implies that a meaningful risk reduction at 10 years also suggests a favorable effect on lifetime risk. Such a study is currently underway and is briefly discussed in the following section.
The ECAD Trial
The ECAD (Eliminate Coronary Artery Disease) trial (NCT02245087) (Figure 1) is a 10-year randomized, controlled, event-driven trial examining whether incident ASCVD events are more effectively prevented by the institution of statin-based LDL-C lowering compared with guideline-based treatment in a population of patients in their early- to mid-middle years who are not yet candidates for guideline-based pharmacological LDL-C lowering, but from which the majority of incident CHD will ultimately come. Thus, the ECAD trial asks whether lowering LDL-C before complex plaques are common in the population will reduce incident CHD compared with guideline-based therapy.
In asymptomatic middle-aged adults with 1 risk factor for developing ASCVD (see inclusion criteria for risk factor list in the following text), addition of atorvastatin 20 mg orally to standard American College of Cardiology/American Heart Association guideline-based treatment does not improve the composite outcome of incident death (due to causes other than cancer or trauma), nonfatal MI, nonfatal stroke, or revascularization.
The primary endpoint was death due to causes other than cancer or trauma, nonfatal MI, coronary revascularization, or nonfatal stroke.
Secondary endpoints included the following:
1. Death (excluding death due to cancer or trauma);
2. Nonfatal MI;
3. Nonfatal stroke;
4. Coronary revascularization;
5. Difference in low-density lipoprotein between usual care and atorvastatin-treated groups;
6. On-treatment (treatment as assigned) analysis of the primary endpoint; and
7. New onset of diabetes.
Particularly interesting among the secondary endpoints will be the on-treatment analysis. The on-treatment (treatment as assigned) analysis will answer the question of the effectiveness of the ECAD management approach when patients are compliant with the treatment strategy. The primary (intention-to-treat) analysis will vet the overall treatment strategy, including compliance with the medication.
Inclusion and exclusion criteria are shown in Table 1. Patients will be recruited for the ECAD trial at sites in Ontario, Canada, and Mexico City, Mexico. The inclusion/exclusion criteria avoid enrolling subjects who should receive therapy on the basis of current U.S. or Canadian guidelines or who are at very low lifetime risk for clinical ASCVD events. Rather, ECAD focuses on the large number of people in the population at low to moderate short-term risk, but at high lifetime risk, for which pharmacological therapy is not now recommended for primary prevention. The question in ECAD is whether the ASCVD clinical event incidence can be dramatically reduced by starting statin therapy in asymptomatic middle-aged adults with at least 1 risk factor, rather than waiting for them to become candidates on the basis of development of either incident symptomatic ASCVD or higher short-term risk that qualifies them for treatment according to current practice guidelines.
Patient flow: randomization
Patients will be randomized 1:1 to receive: 1) atorvastatin 20 mg daily in addition to guideline-based care; or 2) guideline-based care.
All subjects randomized to active therapy will continue taking the study medication (atorvastatin 20 mg daily) until the end of the study (potentially 8 to 10 years, depending on the time of recruitment) unless there is a clinical decision to discontinue medication or the subject decides to withdraw from treatment.
Patients in the active treatment arm will receive atorvastatin 20 mg daily, which is expected to lower the LDL-C level by approximately 43%. Atorvastatin is provided by Pfizer Inc. (New York, New York), distributed to participating sites by the Advanced Medical Research Institute of Canada, and dispensed directly to participating subjects by local clinic staff.
Atorvastatin was selected in the dose chosen because it is a drug with known efficacy and, at a dose of 20 mg daily, is well-tolerated. A higher dose would produce greater LDL-C lowering, but with an increased incidence of side effects that might compromise compliance. No run-in period is contemplated for ECAD.
Patient flow: post-randomization
For subjects randomized to the active arm, a follow-up blood draw will be scheduled 10 weeks post-randomization to perform liver function tests and determine cholesterol levels. Individuals with transaminase levels that rise to ≥3× the upper limit of normal will be instructed to cease taking atorvastatin. Patients with muscular pains will have a serum creatine phosphokinase (CPK) drawn, and the drug will be discontinued if there is a rise to >2× the upper limit of normal for CPK. Atorvastatin will also be stopped if the patient considers the pain intolerable, regardless of CPK level. If the patient experiences an allergic reaction, atorvastatin will be stopped and not restarted.
Subsequent visits for routine clinical care, anticipated to be approximately annually, will be used to assess and encourage compliance (as described later).
Subjects in both arms will continue to receive their routine medical assessments and standard medical care, including guideline-based statin, if indicated. An event-driven type of trial, initially randomizing 10,000 to 15,000 patients, is contemplated, with follow-up continuing until the desired number of total primary endpoint events have accumulated. We calculate that a total of 330 events would be required to yield 90% power for a 30% risk reduction (Online Appendix 1). This number of events would likely be accrued in the ninth year if the cumulative event rate were 3%, but would accrue in the sixth year at the higher end of our anticipated event rate range.
Outcomes in Canada will be assessed using provincial administrative databases housed at the Institute for Clinical Evaluative Sciences. Both MI and coronary revascularization have excellent coding validity and are captured through the Discharge Abstract Database. Mortality data will be captured using the Registered Persons Database. Ontario Health Insurance Plan numbers will be collected at enrollment for future linkages through databases housed at the Institute for Clinical Evaluative Sciences for this purpose. If a coronary revascularization is confirmed, the angiography report will also be collected. At the beginning of the study, all enrolled subjects will be asked to provide consent to collect this information in the event it is needed. Because international sites are contemplated, endpoint ascertainment in these sites will require use of follow-up facilities in those countries.
The primary outcome in this trial is the time of the first occurrence of MI, coronary revascularization, stroke, or death from any cause other than cancer or trauma in the follow-up period. A secondary analysis will consider the components of the composite outcome from a multiple/recurrent events perspective. Specifically, the marginal model of Wei, Lin, and Weissfeld (31) will be considered. On-treatment analysis is also planned. Subgroups pre-specified for analysis include men, women, and subjects of South Asian descent (Online Appendix 1).
Trial operation is governed by a steering committee (Online Appendix 2 for the roster). The study organization will include a data and safety monitoring board (Online Appendix 3 for the data and safety monitoring board roster) charged with reviewing the scientific conduct of the study and making recommendations about whether or not to continue the study, as well as recommendations about any aspect of trial operation. A recommendation to stop the trial for overt efficacy or futility will be largely on the basis of protocol-specified interim analyses incorporating stopping guidelines. Interim analyses of efficacy will be relatively conservative. Annual interim analyses will start at the end of the fifth year, and will initially require p < 0.0001 (years 5 and 6) and subsequently require p < 0.001 (years 7 to 9) to consider recommending early termination of the study.
Primary Prevention After ECAD
The results of the ECAD trial will guide both practice and future research directions. A significant reduction in CHD incidence would be compelling evidence of the therapeutic value of early LDL-C lowering in a large population of patients who are not now being treated. If ECAD shows a significant reduction in incident ASCVD with early LDL-C lowering, in addition to guideline-based therapy, analysis using NHANES (National Health and Nutrition Examination Survey) data (Figure 2) suggests that statin treatment would be indicated in an additional 10 million to 11 million people in the United States alone (personal communication, D.M. Lloyd-Jones, February 2015).
Also, a very large effect size would suggest that LDL-C lowering was started early enough in life to have a major preventive effect and that the degree of LDL-C lowering is adequate. A quantitatively less impressive effect size would guide practice and further research regarding how soon to start pharmacological LDL-C lowering and how great the LDL-C lowering should be to optimize primary prevention.
Although low cholesterol dating from birth or childhood would likely result in the most effective prevention of CHD and other ASCVD clinical events, reduction of LDL-C using a low-dose statin in a population of early- to mid-middle aged subjects, who had complex plaques infrequently and who do not reach a threshold for treatment in the current guidelines, may prove highly effective without the need to expose patients to a lifetime of pharmacological therapy. The ECAD trial is designed to address this question.
Pfizer, Inc., has provided the statin medication for the ECAD trial. Dr. Domanski is the steering committee chairman of the ECAD trial. Dr. Diaz-Mitoma has conducted clinical trials under contract for Merck and GlaxoSmithKline. Dr. Mamdani has served as an advisory board member for AstraZeneca, Bristol-Myers Squibb, Eli Lilly and Company, GlaxoSmithKline, Hoffman La Roche, Novartis, Novo Nordisk, and Pfizer, Inc. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose. John Kastelein, MD, PhD, served as Guest Editor for this paper.
- Abbreviations and Acronyms
- atherosclerotic cardiovascular disease
- coronary heart disease
- creatine phosphokinase
- low-density lipoprotein cholesterol
- myocardial infarction
- Received July 15, 2015.
- Accepted August 18, 2015.
- American College of Cardiology Foundation
- Fuster V.,
- Kelly B.B.,
- Vedanthan R.
- Wikipedia. List of wars and anthropogenic disasters by death toll. 2015. Available at: https://en.wikipedia.org/wiki/List_of_wars_and_anthropogenic_disasters_by_death_toll. Accessed August 19, 2015.
- Heidenreich P.A.,
- Trogdon J.G.,
- Khavjou O.A.,
- et al.,
- for the American Heart Association Advocacy Coordinating Committee,
- Stroke Council,
- Council on Cardiovascular Radiology and Intervention,
- Council on Clinical Cardiology,
- Council on Epidemiology and Prevention,
- Council on Arteriosclerosis, Thrombosis and Vascular Biology,
- Council on Cardiopulmonary Critical Care, Perioperative and Resuscitation,
- Council on Cardiovascular Nursing,
- Council on the Kidney in Cardiovascular Disease,
- Council on Cardiovascular Surgery and Anesthesia,
- Interdisciplinary Council on Quality of Care and Outcomes Research
- Chen Z.,
- Peto R.,
- Collins R.,
- et al.
- Stone N.J.,
- Robinson J.G.,
- Lichtenstein A.H.,
- et al.
- Stamler J.,
- Wentworth D.,
- Neaton J.D.
- Stamler J.
- Fernández-Friera L.,
- Peñalvo J.L.,
- Fernández-Ortiz A.,
- et al.
- Usman B.,
- Mehran R.,
- Sartori S.,
- et al.
- Sever P.S.,
- Dahlöf B.,
- Poulter N.R.,
- et al.,
- for the ASCOT investigators
- Hsia J.,
- MacFadyen J.G.,
- Monyak J.,
- et al.
- Mozaffarian D.,
- Benjamin E.J.,
- Go A.S.,
- et al.,
- for the American Heart Association Statistics Committee and Stroke Statistics Subcommittee
- Vint F.W.
- Kato H.,
- Tillotson J.,
- Nichaman M.Z.,
- et al.
- Truswell A.S.,
- Hansen D.L.
- Law M.R.,
- Wald N.J.,
- Rudnicka A.R.